File size: 149,728 Bytes
6fa4bc9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 |
{
"paper_id": "N15-1039",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T14:33:08.961665Z"
},
"title": "I Can Has Cheezburger? A Nonparanormal Approach to Combining Textual and Visual Information for Predicting and Generating Popular Meme Descriptions",
"authors": [
{
"first": "William",
"middle": [
"Yang"
],
"last": "Wang",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Carnegie Mellon University Pittsburgh",
"location": {
"postCode": "15213",
"region": "PA"
}
},
"email": ""
},
{
"first": "Miaomiao",
"middle": [],
"last": "Wen",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Carnegie Mellon University Pittsburgh",
"location": {
"postCode": "15213",
"region": "PA"
}
},
"email": ""
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "The advent of social media has brought Internet memes, a unique social phenomenon, to the front stage of the Web. Embodied in the form of images with text descriptions, little do we know about the \"language of memes\". In this paper, we statistically study the correlations among popular memes and their wordings, and generate meme descriptions from raw images. To do this, we take a multimodal approach-we propose a robust nonparanormal model to learn the stochastic dependencies among the image, the candidate descriptions, and the popular votes. In experiments, we show that combining text and vision helps identifying popular meme descriptions; that our nonparanormal model is able to learn dense and continuous vision features jointly with sparse and discrete text features in a principled manner, outperforming various competitive baselines; that our system can generate meme descriptions using a simple pipeline.",
"pdf_parse": {
"paper_id": "N15-1039",
"_pdf_hash": "",
"abstract": [
{
"text": "The advent of social media has brought Internet memes, a unique social phenomenon, to the front stage of the Web. Embodied in the form of images with text descriptions, little do we know about the \"language of memes\". In this paper, we statistically study the correlations among popular memes and their wordings, and generate meme descriptions from raw images. To do this, we take a multimodal approach-we propose a robust nonparanormal model to learn the stochastic dependencies among the image, the candidate descriptions, and the popular votes. In experiments, we show that combining text and vision helps identifying popular meme descriptions; that our nonparanormal model is able to learn dense and continuous vision features jointly with sparse and discrete text features in a principled manner, outperforming various competitive baselines; that our system can generate meme descriptions using a simple pipeline.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "In the past few years, Internet memes become a new, contagious social phenomenon: it all starts with an image with a witty, catchy, or sarcastic sentence, and people circulate it from friends to friends, colleagues to colleagues, and families to families. Eventually, some of them go viral on the Internet.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Meme is not only about the funny picture, the Internet culture, or the emotion that passes along, but also about the richness and uniqueness of its language: it is often highly structured with special written style, and forms interesting and subtle connotations that resonate among the readers. For example, the LOL cat memes (e.g., Figure 1 ) often include superimposed text with broken grammars and/or spellings.",
"cite_spans": [],
"ref_spans": [
{
"start": 333,
"end": 341,
"text": "Figure 1",
"ref_id": "FIGREF0"
}
],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Even though the memes are popular over the Internet, the \"language of memes\" is still not wellunderstood: there are no systematic studies on predicting and generating popular Internet memes from the Natural Language Processing (NLP) and Computer Vision (CV) perspectives.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In this paper, we take a multimodal approach to predict and generate popular meme descriptions. To do this, we collect a set of original meme images, a list of candidate descriptions, and the corresponding votes. We propose a robust nonparanormal approach (Liu et al., 2009) to model the multimodal stochastic dependencies among images, text, and votes. We then introduce a simple pipeline for generating meme descriptions combining reverse image search and traditional information retrieval approaches. In empirical experiments, we show that our model outperforms strong discriminative baselines by very large margins in the regression/ranking experiments, and that in the generation experiment, the nonparanormal outperforms the second-best supervised baseline by 4.35 BLEU points, and obtains a BLEU score improvement of 4.48 over an unsupervised recurrent neural network language model trained on a large meme corpus that is almost 90 times larger. Our contributions are three-fold:",
"cite_spans": [
{
"start": 256,
"end": 274,
"text": "(Liu et al., 2009)",
"ref_id": "BIBREF33"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "\u2022 We are the first to study the \"language of memes\" combining NLP, CV, and machine learning techniques, and show that combining the visual and textual signals helps identifying popular meme descriptions;",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "\u2022 Our approach empowers Internet users to select better wordings and generate new memes automatically;",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "\u2022 Our proposed robust nonparanormal model outperforms competitive baselines for predicting and generating popular meme descriptions.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In the next section, we outline related work. In Section 3, we introduce the theory of copula, and our nonparanormal approach. In Section 4, we describe the datasets. We show the prediction and generation results in Section 5 and Section 6. Finally, we conclude in Section 7.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Although the language of Internet memes is a relatively new research topic, our work is broadly related to studies on predicting popular social media messages (Hong et al., 2011; Bakshy et al., 2011; Artzi et al., 2012) . Most recently, Tan et al. (2014) study the effect on wordings for Tweets. However, none of the above studies have investigated multimodal approaches that combine text and vision.",
"cite_spans": [
{
"start": 159,
"end": 178,
"text": "(Hong et al., 2011;",
"ref_id": "BIBREF26"
},
{
"start": 179,
"end": 199,
"text": "Bakshy et al., 2011;",
"ref_id": "BIBREF2"
},
{
"start": 200,
"end": 219,
"text": "Artzi et al., 2012)",
"ref_id": "BIBREF1"
},
{
"start": 237,
"end": 254,
"text": "Tan et al. (2014)",
"ref_id": "BIBREF48"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2"
},
{
"text": "Recently, there has been growing interests in inter-disciplinary research on generating image descriptions. Gupta el al. (2009) have studied the problem of constructing plots from video understanding. The work by Farhadi et al. (2010) is among the first to generate sentences from images. Kulkarni et al. (2011) use linguistic constraints and a conditional random field model for the task, whereas leverage syntactic information and co-occurrence statistics and use a large text corpus and CV algorithms for detecting visual text. With the surge of interests in deep learning techniques in NLP (Socher et al., 2013; Devlin et al., 2014) and CV Oquab et al., 2013) , there have been several unrefereed manuscripts on parsing images and generating text descriptions lately (Vinyals et al., 2014; Chen and Zitnick, 2014; Donahue et al., 2014; Fang et al., 2014; Karpathy and Fei-Fei, 2014) using neural network models. Although the above studies have shown interesting results, our task is arguably more complex than generating text descriptions: in addition to the visual and textual signals, we have to model the popular votes as a third dimension for learning. For example, we cannot simply train a convolutional neural network image parser on billions of images, and use recurrent neural networks to generate texts such as \"There is a white cat sitting next to a laptop.\" for Figure 1. Additionally, since not all images are suitable as meme images, collecting training images is also more challenging in our task.",
"cite_spans": [
{
"start": 108,
"end": 127,
"text": "Gupta el al. (2009)",
"ref_id": null
},
{
"start": 213,
"end": 234,
"text": "Farhadi et al. (2010)",
"ref_id": "BIBREF16"
},
{
"start": 289,
"end": 311,
"text": "Kulkarni et al. (2011)",
"ref_id": "BIBREF31"
},
{
"start": 594,
"end": 615,
"text": "(Socher et al., 2013;",
"ref_id": "BIBREF47"
},
{
"start": 616,
"end": 636,
"text": "Devlin et al., 2014)",
"ref_id": "BIBREF9"
},
{
"start": 644,
"end": 663,
"text": "Oquab et al., 2013)",
"ref_id": "BIBREF39"
},
{
"start": 771,
"end": 793,
"text": "(Vinyals et al., 2014;",
"ref_id": "BIBREF51"
},
{
"start": 794,
"end": 817,
"text": "Chen and Zitnick, 2014;",
"ref_id": "BIBREF6"
},
{
"start": 818,
"end": 839,
"text": "Donahue et al., 2014;",
"ref_id": "BIBREF12"
},
{
"start": 840,
"end": 858,
"text": "Fang et al., 2014;",
"ref_id": "BIBREF15"
},
{
"start": 859,
"end": 886,
"text": "Karpathy and Fei-Fei, 2014)",
"ref_id": "BIBREF28"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2"
},
{
"text": "In contrast to prior work, we take a very different approach: we investigate copula methods (Schweizer and Sklar, 1983; Nelsen, 1999) , in particular, the nonparanormals (Liu et al., 2009) , for joint modeling of raw images, text descriptions, and popular votes. Copula is a statistical framework for analyzing random variables from Statistics , and often used in Economics (Chen and Fan, 2006) . Only until very recently, researchers from the machine learning and information retrieval communities (Ghahramani et al., 2012; Han et al., 2012; Eickhoff et al., 2013) . start to understand the theory and the predictive power of copula models. Wang and Hua (2014) are the first to introduce semiparametric Gaussian copula (a.k.a. nonparanormals) for text prediction. However, their approach may be prone to overfitting. In this work, we generalize Wang and Hua's method to jointly model text and vision features with popular votes, while scaling up the model using effective dropout regularization.",
"cite_spans": [
{
"start": 92,
"end": 119,
"text": "(Schweizer and Sklar, 1983;",
"ref_id": "BIBREF44"
},
{
"start": 120,
"end": 133,
"text": "Nelsen, 1999)",
"ref_id": "BIBREF37"
},
{
"start": 170,
"end": 188,
"text": "(Liu et al., 2009)",
"ref_id": "BIBREF33"
},
{
"start": 374,
"end": 394,
"text": "(Chen and Fan, 2006)",
"ref_id": "BIBREF5"
},
{
"start": 499,
"end": 524,
"text": "(Ghahramani et al., 2012;",
"ref_id": "BIBREF20"
},
{
"start": 525,
"end": 542,
"text": "Han et al., 2012;",
"ref_id": "BIBREF34"
},
{
"start": 543,
"end": 565,
"text": "Eickhoff et al., 2013)",
"ref_id": "BIBREF13"
},
{
"start": 642,
"end": 661,
"text": "Wang and Hua (2014)",
"ref_id": "BIBREF52"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2"
},
{
"text": "A key challenge for joint modeling of text and vision is that, because textual features are often relatively sparse and discrete, while visual features are typically dense and continuous, it is difficult to model them jointly in a principled way.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Our Approach",
"sec_num": "3"
},
{
"text": "To avoid comparing \"apple and oranges\" in the same probabilistic space, we propose the nonparanormal approach, which extends the Gaussian graphical model by transforming its variables by smooth functions. More specifically, for each dimension of textual and visual features, instead of using raw counts or histograms, we first use probability integral transform to generate empirical cumulative density functions (ECDF): now instead of the probability density function (PDF) space, we are working in the ECDF space where the value of each feature is based on the rank, and is strictly restricted between 0 and 1. Then, we use kernel density estimation to smooth out the zeroing features 1 . Finally, now textual and visual features are compatible, and we then build a parametric Gaussian copula model to estimate the pair-wise correlations among the covariate and the dependent variable.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Our Approach",
"sec_num": "3"
},
{
"text": "In this section, we first explain the visual and textual features used in this study. Then, we introduce the theory of copula, and describe the robust nonparanormal. Finally, we show a simple pipeline for generating meme descriptions.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Our Approach",
"sec_num": "3"
},
{
"text": "Textual Features To model the meme descriptions, we take a broad range of textual features into considerations:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "\u2022 Lexical Features: we extract unigrams and bigrams from meme descriptions as surface-level lexical features.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "\u2022 Part-of-Speech Features: to model shallow syntactic cues, we extract lexicalized part-ofspeech features using the Stanford part-ofspeech tagger (Toutanova et al., 2003) .",
"cite_spans": [
{
"start": 146,
"end": 170,
"text": "(Toutanova et al., 2003)",
"ref_id": "BIBREF50"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "\u2022 Dependency Triples: to better understand the deeper syntactic dependencies of keywords in Figure 3 : An example of the standard SIFT keypoints detected on the \"doge\" meme.",
"cite_spans": [],
"ref_spans": [
{
"start": 92,
"end": 100,
"text": "Figure 3",
"ref_id": null
}
],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "memes, we have also extracted typed dependency triples (e.g., subj(I,are)) using the Malt-Parser (Nivre et al., 2007) .",
"cite_spans": [
{
"start": 97,
"end": 117,
"text": "(Nivre et al., 2007)",
"ref_id": "BIBREF38"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "\u2022 Named Entity Features: after browsing the dataset, we notice that certain names are often mentioned in memes (e.g. \"Drake\", \"Kenye West\", and \"Justin Bieber\"), so we utilize the Stanford named entity recognizer (Finkel et al., 2005) to extract lexicalized named entities.",
"cite_spans": [
{
"start": 213,
"end": 234,
"text": "(Finkel et al., 2005)",
"ref_id": "BIBREF18"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "\u2022 Frame-Semantics Features: SEMAFOR (Das et al., 2010 ) is a state-of-the-art framesemantics parser that produces FrameNet-style semantic annotation. We use SEMAFOR to extract frame-level semantic features.",
"cite_spans": [
{
"start": 36,
"end": 53,
"text": "(Das et al., 2010",
"ref_id": "BIBREF8"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "Visual Features A key insight on viral memes is that the images producing a shared social signal are typically inter-related in style. For example, LOLcats are an early series of memes involving funny cat photos. Similarly, \"Bieber memes\" involve modified pictures of Bieber. Therefore, we hypothesize that, by extracting visual features, it is of crucial importance to capture the entities, objects, and styles as visual words in these inter-related meme images. The popular visual bag-of-words representation (Sivic and Zisserman, 2003 ) is used to describe images:",
"cite_spans": [
{
"start": 511,
"end": 537,
"text": "(Sivic and Zisserman, 2003",
"ref_id": "BIBREF45"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "1. PHOW Features Extraction: unlike text features, SIFT first detects the Harris keypoints from an image, and then describes each keypoint with a vector. An example of the SIFT frames are shown in Figure 3 . PHOW (Bosch et al., 2007) is a dense and multi-scale variant of the Scale Invariant Feature Transform (SIFT) descriptors. Using PHOW, we obtain about 20K keypoints for each image.",
"cite_spans": [
{
"start": 213,
"end": 233,
"text": "(Bosch et al., 2007)",
"ref_id": "BIBREF3"
}
],
"ref_spans": [
{
"start": 197,
"end": 205,
"text": "Figure 3",
"ref_id": null
}
],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "2. Elkan K-means Clustering is the clustering method (Elkan, 2003) that we use to obtain the vocabulary for visual words. Comparing to other variants of K-means, this method quickly constructs the codebook from PHOW keypoints. 3. Bag-of-Words Histograms are used to represent each image. We match the PHOW keypoints of each image with the vocabulary that we extract from the previous step, and generate a 1 \u00d7 200 sized visual bag-of-words vector.",
"cite_spans": [
{
"start": 53,
"end": 66,
"text": "(Elkan, 2003)",
"ref_id": "BIBREF14"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Features",
"sec_num": "3.1"
},
{
"text": "In the Statistics literature, copula is widely known as a family of distribution function. The idea behind copula theory is that the cumulative distribution function (CDF) of a random vector can be represented in the form of uniform marginal cumulative distribution functions, and a copula that connects these marginal CDFs, which describes the correlations among the input random variables. However, in order to have a valid multivariate distribution function regardless of n-dimensional covariates, not every function can be used as a copula function. The central idea behind copula, therefore, can be summarize by the Sklar's theorem and the corollary. Theorem 1 (Sklar's Theorem (1959) ) Let F be the joint cumulative distribution function of n random variables X 1 , X 2 , ..., X n . Let the corresponding marginal cumulative distribution functions of the random variable be",
"cite_spans": [
{
"start": 666,
"end": 689,
"text": "(Sklar's Theorem (1959)",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "The Theory of Copula",
"sec_num": "3.2"
},
{
"text": "F 1 (x 1 ), F 2 (x 2 ), ..., F n (x n ).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Theory of Copula",
"sec_num": "3.2"
},
{
"text": "Then, if the marginal functions are continuous, there exists a unique copula C, such that",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Theory of Copula",
"sec_num": "3.2"
},
{
"text": "F (x 1 , ..., x n ) = C[F 1 (x 1 ), ..., F n (x n )]. (1)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Theory of Copula",
"sec_num": "3.2"
},
{
"text": "Furthermore, if the distributions are continuous, the multivariate dependency structure and the marginals might be separated, and the copula can be considered independent of the marginals (Joe, 1997; Parsa and Klugman, 2011) . Therefore, the copula does not have requirements on the marginal distributions, and any arbitrary marginals can be combined and their dependency structure can be modeled using the copula. The inverse of Sklar's Theorem is also true in the following: Corollary 1 If there exists a copula C : (0, 1) n and marginal cumulative distribution functions",
"cite_spans": [
{
"start": 188,
"end": 199,
"text": "(Joe, 1997;",
"ref_id": "BIBREF27"
},
{
"start": 200,
"end": 224,
"text": "Parsa and Klugman, 2011)",
"ref_id": "BIBREF41"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "The Theory of Copula",
"sec_num": "3.2"
},
{
"text": "F 1 (x 1 ), F 2 (x 2 ), ..., F n (x n ), then C[F 1 (x 1 ), ..., F n (x n )",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Theory of Copula",
"sec_num": "3.2"
},
{
"text": "] defines a multivariate cumulative distribution function.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Theory of Copula",
"sec_num": "3.2"
},
{
"text": "To model multivariate text and vision variables, we choose the nonparanormal (NPN) as the copula function in this study, which can be explained in the following two parts.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparanormal",
"sec_num": "3.3"
},
{
"text": "Assume we have n random variables of vision and text features X 1 , X 2 , ..., X n . The problem is that text features are sparse, so we need to perform nonparametric kernel density estimation to smooth out the distribution of each variable. Let f 1 , f 2 , ..., f n be the unknown density, we are interested in deriving the shape of these functions. Assume we have m samples, the kernel density estimator can be defined as:f",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "h (x) = 1 m m i=1 K h (x \u2212 x i ) (2) = 1 mh m i=1 K x \u2212 x i h (3)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "Here, K(\u2022) is the kernel function, where in our case, we use the Box kernel 2 K(z):",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "K(z) = 1 2 , |z| \u2264 1,",
"eq_num": "(4)"
}
],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "= 0, |z| > 1.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "Comparing to the Gaussian kernel and other kernels, the Box kernel is simple, and computationally inexpensive. The parameter h is the bandwidth for smoothing 3 . Now, we can derive the empirical cumulative distribution function\u015d",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "F X 1 (f 1 (X 1 )),F X 2 (f 2 (X 2 )), ...,F Xn (f n (X n ))",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "of the smoothed covariates, as well as the dependent variable y (which is the reciprocal rank of the popular votes of a meme) and its CDFF y (f (y)). The empirical cumulative distribution functions are defined as:F",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "(\u03bd) = 1 m m i=1 I{x i \u2264 \u03bd} (6)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "where I{\u2022} is the indicator function, and \u03bd indicates the current value that we are evaluating. Note that the above step is also known as probability integral transform (Diebold et al., 1997) , which allows us to convert any given continuous distribution to random variables having a uniform distribution. This is crucial for text: instead of using the raw counts, we are now working with uniform marginal CDFs, which helps coping with the overfitting issue due to noise and data sparsity. We also use the same procedure to transform the vision features into CDF space to be compatible with text features.",
"cite_spans": [
{
"start": 169,
"end": 191,
"text": "(Diebold et al., 1997)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "The Robust Estimation of Copula Now that we have obtained the marginals, and then the joint distribution can be constructed by applying the copula function that models the stochastic dependencies among marginal CDFs:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "F (f 1 (X 1 ), ...,f 1 (X n ),f (y)) = C[F X1 f 1 (X 1 ) , ...,F Xn f n (X n ) ,F y f y (y) ]",
"eq_num": "(7)"
}
],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "In this work, we apply the parametric Gaussian copula to model the correlations among the text features and the label. Assume x i is the smoothed version of random variable X i , and y is the smoothed label, we have:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "F (x 1 , ..., x n , y) = \u03a6 \u03a3 \u03a6 \u22121 [F x1 (x 1 )], ..., , \u03a6 \u22121 [F xn (x n )], \u03a6 \u22121 [F y (y)]",
"eq_num": "(8)"
}
],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "where \u03a6 \u03a3 is the joint cumulative distribution function of a multivariate Gaussian with zero mean and \u03a3 variance. \u03a6 \u22121 is the inverse CDF of a standard Gaussian. In this parametric part of the model, the parameter estimation boils down to the problem of learning the covariance matrix \u03a3 of this Gaussian copula. In this work, we perform standard maximum likelihood estimation (MLE) for the \u03a3 matrix, where we follow the details from prior work (Wang and Hua, 2014) . To avoid overfitting, traditionally, one resorts to classic regularization techniques such as Lasso (Tib-shirani, 1996) . While Lasso is widely used, the nondifferentiable nature of the L 1 norm often make the objective function difficult to optimize. In this work, we propose dropout training as copula regularization. Dropout was proposed by Hinton et al. as a method to prevent feature coadaptation in the deep learning framework, but recently studies (Wager et al., 2013) also show that its behaviour is similar to L 2 regularization, and can be approximated efficiently (Wang and Manning, 2013) in many other machine learning tasks. Another advantage of dropout training is that, unlike Lasso, it does not require all the features for training, and training is \"embarrassingly\" parallelizable.",
"cite_spans": [
{
"start": 444,
"end": 464,
"text": "(Wang and Hua, 2014)",
"ref_id": "BIBREF52"
},
{
"start": 567,
"end": 586,
"text": "(Tib-shirani, 1996)",
"ref_id": null
},
{
"start": 1042,
"end": 1066,
"text": "(Wang and Manning, 2013)",
"ref_id": "BIBREF53"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "In Gaussian copula estimation context, we can introduce another dimension : the number of dropout learners, to extend the \u03a3 into a dropout tensor. Essentially, the task becomes the estimation of \u03a3 1 , \u03a3 2 , ..., \u03a3 where the input feature space for each dropout component is randomly corrupted by (1 \u2212 \u03b4) percent of the original dimension. In the inference time, we use geometric mean to average the predictions from each dropout learner, and generate the final prediction. Note that the final \u03a3 matrix has to be symmetric and positive definite, so we apply tiny random Gaussian noise to maintain the property.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Nonparametric Estimation",
"sec_num": null
},
{
"text": "One important question regarding the proposed nonparanormal model is the corresponding computational complexity. This boils down to the estimation of the\u03a3 matrix : one only needs to calculate the correlation coefficients of n(n \u2212 1)/2 pairs of random variables. Christensen (2005) shows that sorting and balanced binary trees can be used to calculate the correlation coefficients with complexity of O(n log n). Therefore, the computational complexity of MLE for the proposed model is O(n log n).",
"cite_spans": [
{
"start": 262,
"end": 280,
"text": "Christensen (2005)",
"ref_id": "BIBREF7"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Computational Complexity",
"sec_num": null
},
{
"text": "In this prediction task, in order to perform the exact inference of the conditional probability distribution p(F y (y)|F x 1 (x 1 ), ..., F xn (x n )), one needs to solve the mean respons\u00ea E(F y (y)|F x 1 (x 1 ), ..., F x 1 (x 1 )) from a joint distribution of high-dimensional Gaussian copula. Unfortunately, the exact inference can be intractable in the multivariate case, and approximate inference, such as Markov Chain Monte Carlo sampling (Gelfand and Smith, 1990; Pitt et al., 2006) is often used for posterior inference. In this work, we propose an efficient sampling method to derive y given the text features -we sampl\u00ea F y (y) s.t. it maximizes the joint high-dimensional Gaussian copula density:",
"cite_spans": [
{
"start": 444,
"end": 469,
"text": "(Gelfand and Smith, 1990;",
"ref_id": "BIBREF19"
},
{
"start": 470,
"end": 488,
"text": "Pitt et al., 2006)",
"ref_id": "BIBREF43"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Efficient Approximate Inference",
"sec_num": null
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "arg max Fy(y)\u2208(0,1) 1 \u221a det \u03a3 exp \u2212 1 2 \u2206 T \u2022 \u03a3 \u22121 \u2212 I \u2022 \u2206",
"eq_num": "(9)"
}
],
"section": "Efficient Approximate Inference",
"sec_num": null
},
{
"text": "where",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Efficient Approximate Inference",
"sec_num": null
},
{
"text": "\u2206 = \uf8eb \uf8ec \uf8ec \uf8ec \uf8ed \u03a6 \u22121 (F x 1 (x 1 )) . . . \u03a6 \u22121 (F xn (x n )) \u03a6 \u22121 (F y (y)) \uf8f6 \uf8f7 \uf8f7 \uf8f7 \uf8f8",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Efficient Approximate Inference",
"sec_num": null
},
{
"text": "This approximate inference scheme using maximum density sampling from the Gaussian copula significantly relaxes the complexity of inference. Finally, to derive\u0177, the last step is to compute the inverse CDF ofF y (y). A detailed description of the inference algorithm can be found in our prior work (Wang and Hua, 2014) .",
"cite_spans": [
{
"start": 298,
"end": 318,
"text": "(Wang and Hua, 2014)",
"ref_id": "BIBREF52"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Efficient Approximate Inference",
"sec_num": null
},
{
"text": "Now after we train a nonparanormal model for ranking meme descriptions, we show the simple meme generation pipeline in Figure 4 .",
"cite_spans": [],
"ref_spans": [
{
"start": 119,
"end": 127,
"text": "Figure 4",
"ref_id": "FIGREF2"
}
],
"eq_spans": [],
"section": "A Simple Meme Generation Pipeline",
"sec_num": "3.4"
},
{
"text": "Given a test image, we disguise as the Internet Explorer, and query Google's \"Search By Image\" inverse image search service 4 . By comparing the query image with all possible images with their captions in Google's database, a \"Best Guess\" of the keywords in the image is then revealed.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A Simple Meme Generation Pipeline",
"sec_num": "3.4"
},
{
"text": "Using the extracted image keywords, we further query a TF-IDF based Lucene 5 meme search engine, which we indexed with a large number of Webcrawled meme descriptions. After we obtain the candidate generations, we then extract all the text and vision features that we described in Section 3.1. Finally, our nonparanormal model ranks all possible candidates, and selects the final generation with the highest posterior.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "A Simple Meme Generation Pipeline",
"sec_num": "3.4"
},
{
"text": "We collected meme images and text descriptions 6 from two popular meme websites 7 . In the prediction experiment, we use 3,008 image-description pairs for training, and 526 image-description pairs for testing. In the generation experiment, we use 269,473 meme descriptions to index the meme search engine, and 50 randomly selected images for testing. During training, we convert the raw counts of popular votes into reciprocal ranks (e.g., the most popular text descriptions will all have a reciprocal rank of 1, and n-th popular one will have a score of 1/n).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Datasets",
"sec_num": "4"
},
{
"text": "In the first experiment, we compare the proposed NPN with various baselines in a prediction task, since prior literature (Hodosh et al., 2013 ) also suggests using ranking based evaluation for associating images with text descriptions. Throughout the experiment sections, we set = 10, and \u03b4 = 80 as the dropout hyperparameters.",
"cite_spans": [
{
"start": 121,
"end": 141,
"text": "(Hodosh et al., 2013",
"ref_id": "BIBREF24"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Prediction Experiments",
"sec_num": "5"
},
{
"text": "The baselines are standard squared-loss linear regression, linear kernel SVM, and non-linear (Gaussian) kernel SVM. In a recent empirical study (Fern\u00e1ndez-Delgado et al., 2014 ) that evaluates 179 classifiers from 17 families on 121 UCI datasets, the authors find that Gaussian SVM is one of the top performing classifiers. We use the Statistical Toolbox's linear regression implementation in Matlab, and LibSVM (Chang and Lin, 2011) for training and testing the SVM models. The hyperparameter C in linear SVM, and the \u03b3 and C hyperparameters in Gaussian SVM are tuned on the training set using 10-fold cross-validation.",
"cite_spans": [
{
"start": 144,
"end": 175,
"text": "(Fern\u00e1ndez-Delgado et al., 2014",
"ref_id": "BIBREF17"
},
{
"start": 412,
"end": 433,
"text": "(Chang and Lin, 2011)",
"ref_id": "BIBREF4"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Baselines:",
"sec_num": null
},
{
"text": "Spearman's correlation (Hogg and Craig, 1994) and Kendall's tau (Kendall, 1938) have been widely used in many real-valued prediction (regression) problems in NLP (Albrecht and Hwa, 2007; Yogatama et al., 2011) , and here we use them to measure the quality of predicted values\u0177 by comparing to the vector of ground truth y. Kendall's tau is a nonparametric statistical metric that have shown to be inexpensive, robust, and representation independent (Lapata, 2006) . We use paired two-tailed t-test to measure the statistical significance.",
"cite_spans": [
{
"start": 23,
"end": 45,
"text": "(Hogg and Craig, 1994)",
"ref_id": "BIBREF25"
},
{
"start": 50,
"end": 79,
"text": "Kendall's tau (Kendall, 1938)",
"ref_id": null
},
{
"start": 162,
"end": 186,
"text": "(Albrecht and Hwa, 2007;",
"ref_id": "BIBREF0"
},
{
"start": 187,
"end": 209,
"text": "Yogatama et al., 2011)",
"ref_id": "BIBREF54"
},
{
"start": 449,
"end": 463,
"text": "(Lapata, 2006)",
"ref_id": "BIBREF32"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Evaluation Metrics:",
"sec_num": null
},
{
"text": "The first two figures in Figure 5 show the learning curve of our system, comparing other baselines. We see that when increasing the amount of training data, our approach clearly dominates all other methods by a large margin. Linear and Gaussian SVMs perform similarly, and have good performances with only 25% of the training data, but the improvements are not large when increasing the amount of training data.",
"cite_spans": [],
"ref_spans": [
{
"start": 25,
"end": 33,
"text": "Figure 5",
"ref_id": "FIGREF3"
}
],
"eq_spans": [],
"section": "Comparison with Various Baselines",
"sec_num": "5.1"
},
{
"text": "In the last two figures in Figure 5 , we increase the amount of features, and compare various models. We see that the linear regression model overfits with 600 features, and Gaussian SVM outperforms the linear SVM. We see that our NPN model clearly outperforms all baselines by a big gap, and does not overfit.",
"cite_spans": [],
"ref_spans": [
{
"start": 27,
"end": 35,
"text": "Figure 5",
"ref_id": "FIGREF3"
}
],
"eq_spans": [],
"section": "Comparison with Various Baselines",
"sec_num": "5.1"
},
{
"text": "In Table 1 , we systematically compare the contributions of each feature set. First, we see that bigram features clearly improve the performance on top of unigram features. Second, named entities are crucial for further boosting the performance. Third, adding the shallow part-of-speech features does not benefit all models, but the dependency triples are shown to be useful for all methods. Finally, we see that using semantic features helps increasing the performances for most of the cases, and combining text and vision features in our NPN framework doubles the perfor- mance for associating popular votes, meme images, and text descriptions.",
"cite_spans": [],
"ref_spans": [
{
"start": 3,
"end": 10,
"text": "Table 1",
"ref_id": null
}
],
"eq_spans": [],
"section": "Combination of Text and Vision",
"sec_num": "5.2"
},
{
"text": "As we mentioned before, because NPNs model the complex network of random variables, a key issue for training NPN is to prevent the model from overfitting to the training data. So far, none of the prior work have investigated dropout training for regularizing the nonparanormals or even copula in general.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Effects of Dropout Training for Nonparanormals",
"sec_num": "5.3"
},
{
"text": "To empirical test the effects of dropout training for nonparanormals, in addition to our datasets, we also compare with the unregularized copula from Wang and Hua (2014) on predicting financial risks from earnings calls. Table 2 clearly suggests that dropout training for NPNs significant improves the performances on various datasets. Table 3 shows the top ranked text features that are highly correlated with popular votes. We see that the named entity features are useful: Paul Walker, UPS, Bruce Willis, Pencil Guy, Amy Winehouse are recognized as entities in the meme dataset. Dependency triples, as a less-understood feature set, also perform well in this task. For example, xcomp(tell,mean) captures the dependency relation of the popular meme series \"You mean to tell me...\". Interestingly, the transitional dependency feature dep(when,but) plays an important role in the language of memes. The object of a preposition, such as pobj (vegas,in) and pobj(life,of), also made the list. Bigrams are shown to be important features as usual. For example, \"Yo daw\" is a popular meme based on rapper Xzibit's famous reality car show \"Pimp My Ride\", where the rapper customizes people's car according to personal preferences. This viral meme follows the pattern 8 of \"Yo daw(g), I herd you like X (noun), so I put an X in your Y (noun) so you can W (verb) while you Z (verb).\"",
"cite_spans": [
{
"start": 941,
"end": 951,
"text": "(vegas,in)",
"ref_id": null
}
],
"ref_spans": [
{
"start": 221,
"end": 228,
"text": "Table 2",
"ref_id": "TABREF1"
},
{
"start": 336,
"end": 343,
"text": "Table 3",
"ref_id": "TABREF3"
}
],
"eq_spans": [],
"section": "The Effects of Dropout Training for Nonparanormals",
"sec_num": "5.3"
},
{
"text": "The use of pronouns, captured by frame semantics features, is associated with popular memes. We hypothesize that by using pronouns such as \"i\", \"you\", \"we\", and \"they\", the meme recalls personal experiences and emotions, thus connects better with the audience. Finally, we see that the punctuation bigram \"... :\" is an important feature in the language of memes, and Web dialect such as \"y\" (why) also exhibits high correlation with the popular votes.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Qualitative Analysis",
"sec_num": "5.4"
},
{
"text": "In this section, we investigate the performance of our meme generation system using 50 test meme images. To quantitatively evaluate our system, we compare with both unsupervised and supervised baselines. For the unsupervised baselines, we compare with a compact recurrent neural network language model (RNNLM) (Mikolov, 2012) trained on the 3,008 text descriptions of our meme training set, as well as a full model of RNNLM trained on a large meme corpus of 269K sentences 9 . For the supervised baselines, all models are trained on the 3,008 training image-description pairs with labels. All these models can be viewed as different re-ranking methods for the retrieved candidate descriptions. We use BLEU score (Papineni et al., 2002) as the evaluation metric, since the generation task can be viewed as translating raw images into sentences, and it is Figure 6 : Examples from the meme generation experiment. First row: the chemistry cat meme. Second row: the forever alone meme. Third row: the Batman slaps Robin meme. Left column: human generated topvoted meme descriptions on memegenerator.net at the time of writing. Middle column: generated output from RNNLM. Right column: generated output from NPNs. used in many caption generation studies (Vinyals et al., 2014; Chen and Zitnick, 2014; Donahue et al., 2014; Fang et al., 2014; Karpathy and Fei-Fei, 2014) . The generation result is shown in Table 4 . Note that when combining B-1 to B-4 scores, BLEU includes a brevity penalty as described in the original BLEU paper. We see that our NPN model outperforms the best supervised baseline by 4.35 BLEU points, while also obtaining an advantage of 4.48 BLEU points over the full RNNLM, which is trained on a corpus that is \u223c90 times larger, in an unsupervised fashion. When breaking down the results, we see that our NPN's advantage is on generating longer phrases, typically trigrams and four-grams, comparing to the other models. This is very interesting, because generating high-quality long phrases is difficult, since the memes are often short. We show some generation examples in Figure 6 . We see that on the left column, the reference memes are the ones with top votes by the crowd. The first chemistry cat meme includes puns, the second forever alone meme includes reference to the life simulation video game, while the last Batman meme has interesting conversations. In the second column, we see that the memes generated by the full RNNLM model are short, which corresponds to the quantitative results in Table 4 . In the third column, our NPN meme generator was able to generate longer descriptions. Interestingly, it also creates a pun for the chemistry cat meme. Our generation on the forever alone meme is also accurate. In the Batman example, we show that the NPN model makes a sentence-image-mismatch type of error: although the generated sentence includes the entities Batman and Robin, as well as their slapping activity, it was originally created for the \"overly attached girlfriend\" meme 10 .",
"cite_spans": [
{
"start": 310,
"end": 325,
"text": "(Mikolov, 2012)",
"ref_id": "BIBREF35"
},
{
"start": 712,
"end": 735,
"text": "(Papineni et al., 2002)",
"ref_id": "BIBREF40"
},
{
"start": 1249,
"end": 1271,
"text": "(Vinyals et al., 2014;",
"ref_id": "BIBREF51"
},
{
"start": 1272,
"end": 1295,
"text": "Chen and Zitnick, 2014;",
"ref_id": "BIBREF6"
},
{
"start": 1296,
"end": 1317,
"text": "Donahue et al., 2014;",
"ref_id": "BIBREF12"
},
{
"start": 1318,
"end": 1336,
"text": "Fang et al., 2014;",
"ref_id": "BIBREF15"
},
{
"start": 1337,
"end": 1364,
"text": "Karpathy and Fei-Fei, 2014)",
"ref_id": "BIBREF28"
}
],
"ref_spans": [
{
"start": 854,
"end": 862,
"text": "Figure 6",
"ref_id": null
},
{
"start": 1401,
"end": 1408,
"text": "Table 4",
"ref_id": "TABREF5"
},
{
"start": 2091,
"end": 2099,
"text": "Figure 6",
"ref_id": null
},
{
"start": 2520,
"end": 2527,
"text": "Table 4",
"ref_id": "TABREF5"
}
],
"eq_spans": [],
"section": "Generation Experiments",
"sec_num": "6"
},
{
"text": "In this paper, we study the language of memes by jointly learning the image, the description, and the popular votes. In particular, we propose a robust nonparanormal approach to transform all vision and text features into the cumulative density function space. By learning the stochastic dependencies, we show that our model significantly outperforms various competitive baselines in the prediction experiments. In addition, we also propose a simple pipeline for generating memes from raw images, drawing the wisdom from reverse image search and traditional information retrieval perspectives. Finally, we show that our model obtains significant BLEU point improvements over an unsupervised RNNLM baseline trained on a larger corpus, as well as other strong supervised baselines.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusions",
"sec_num": "7"
},
{
"text": "This is necessary for the normal inversion of the ECDFs, which we will describe in Section 3.2.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "It is also known as the original Parzen windows(Parzen, 1962).3 In our implementation, we use the default h of the Box kernel in the ksdensity function in Matlab.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "http://www.google.com/imghp/",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "http://lucene.apache.org/ 6 http://www.cs.cmu.edu/\u02dcyww/data/meme dataset.zip. 7 memegenerator.net and cheezburger.com",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "Note that there are no image features feeding to the unsupervised RNN models.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "http://www.overlyattachedgirlfriend.com",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
}
],
"back_matter": [],
"bib_entries": {
"BIBREF0": {
"ref_id": "b0",
"title": "Regression for sentence-level mt evaluation with pseudo references",
"authors": [
{
"first": "Joshua",
"middle": [],
"last": "Albrecht",
"suffix": ""
},
{
"first": "Rebecca",
"middle": [],
"last": "Hwa",
"suffix": ""
}
],
"year": 2007,
"venue": "Proceedings of ACL",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Joshua Albrecht and Rebecca Hwa. 2007. Regression for sentence-level mt evaluation with pseudo references. In Proceedings of ACL.",
"links": null
},
"BIBREF1": {
"ref_id": "b1",
"title": "Predicting responses to microblog posts",
"authors": [
{
"first": "Yoav",
"middle": [],
"last": "Artzi",
"suffix": ""
},
{
"first": "Patrick",
"middle": [],
"last": "Pantel",
"suffix": ""
},
{
"first": "Michael",
"middle": [],
"last": "Gamon",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of NAACL-HLT",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Yoav Artzi, Patrick Pantel, and Michael Gamon. 2012. Predicting responses to microblog posts. In Proceed- ings of NAACL-HLT.",
"links": null
},
"BIBREF2": {
"ref_id": "b2",
"title": "Everyone's an influencer: quantifying influence on twitter",
"authors": [
{
"first": "Eytan",
"middle": [],
"last": "Bakshy",
"suffix": ""
},
{
"first": "Jake",
"middle": [
"M"
],
"last": "Hofman",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "Winter",
"suffix": ""
},
{
"first": "Duncan J",
"middle": [],
"last": "Mason",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Watts",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of WSDM",
"volume": "",
"issue": "",
"pages": "65--74",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Eytan Bakshy, Jake M Hofman, Winter A Mason, and Duncan J Watts. 2011. Everyone's an influencer: quantifying influence on twitter. In Proceedings of WSDM, pages 65-74. ACM.",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "Image classification using random forests and ferns",
"authors": [
{
"first": "Anna",
"middle": [],
"last": "Bosch",
"suffix": ""
},
{
"first": "Andrew",
"middle": [],
"last": "Zisserman",
"suffix": ""
},
{
"first": "Xavier",
"middle": [],
"last": "Munoz",
"suffix": ""
}
],
"year": 2007,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Anna Bosch, Andrew Zisserman, and Xavier Munoz. 2007. Image classification using random forests and ferns.",
"links": null
},
"BIBREF4": {
"ref_id": "b4",
"title": "Libsvm: a library for support vector machines",
"authors": [
{
"first": "Chih-Chung",
"middle": [],
"last": "Chang",
"suffix": ""
},
{
"first": "Chih-Jen",
"middle": [],
"last": "Lin",
"suffix": ""
}
],
"year": 2011,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Chih-Chung Chang and Chih-Jen Lin. 2011. Libsvm: a library for support vector machines. ACM TIST.",
"links": null
},
"BIBREF5": {
"ref_id": "b5",
"title": "Estimation of copula-based semiparametric time series models",
"authors": [
{
"first": "Xiaohong",
"middle": [],
"last": "Chen",
"suffix": ""
},
{
"first": "Yanqin",
"middle": [],
"last": "Fan",
"suffix": ""
}
],
"year": 2006,
"venue": "Journal of Econometrics",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Xiaohong Chen and Yanqin Fan. 2006. Estimation of copula-based semiparametric time series models. Journal of Econometrics.",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "Learning a recurrent visual representation for image caption generation",
"authors": [
{
"first": "Xinlei",
"middle": [],
"last": "Chen",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Lawrence Zitnick",
"suffix": ""
}
],
"year": 2014,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {
"arXiv": [
"arXiv:1411.5654"
]
},
"num": null,
"urls": [],
"raw_text": "Xinlei Chen and C Lawrence Zitnick. 2014. Learning a recurrent visual representation for image caption gen- eration. arXiv preprint arXiv:1411.5654.",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "Fast algorithms for the calculation of kendalls \u03c4",
"authors": [
{
"first": "David",
"middle": [],
"last": "Christensen",
"suffix": ""
}
],
"year": 2005,
"venue": "Computational Statistics",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "David Christensen. 2005. Fast algorithms for the calcu- lation of kendalls \u03c4 . Computational Statistics.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "Probabilistic frame-semantic parsing",
"authors": [
{
"first": "Dipanjan",
"middle": [],
"last": "Das",
"suffix": ""
},
{
"first": "Nathan",
"middle": [],
"last": "Schneider",
"suffix": ""
},
{
"first": "Desai",
"middle": [],
"last": "Chen",
"suffix": ""
},
{
"first": "Noah A",
"middle": [],
"last": "Smith",
"suffix": ""
}
],
"year": 2010,
"venue": "Proceedings of NAACL-HLT",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Dipanjan Das, Nathan Schneider, Desai Chen, and Noah A Smith. 2010. Probabilistic frame-semantic parsing. In Proceedings of NAACL-HLT.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "Fast and robust neural network joint models for statistical machine translation",
"authors": [
{
"first": "Jacob",
"middle": [],
"last": "Devlin",
"suffix": ""
},
{
"first": "Rabih",
"middle": [],
"last": "Zbib",
"suffix": ""
},
{
"first": "Zhongqiang",
"middle": [],
"last": "Huang",
"suffix": ""
},
{
"first": "Thomas",
"middle": [],
"last": "Lamar",
"suffix": ""
},
{
"first": "Richard",
"middle": [],
"last": "Schwartz",
"suffix": ""
},
{
"first": "John",
"middle": [],
"last": "Makhoul",
"suffix": ""
}
],
"year": 2014,
"venue": "Proceedings of ACL",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Jacob Devlin, Rabih Zbib, Zhongqiang Huang, Thomas Lamar, Richard Schwartz, and John Makhoul. 2014. Fast and robust neural network joint models for statis- tical machine translation. In Proceedings of ACL.",
"links": null
},
"BIBREF10": {
"ref_id": "b10",
"title": "Evaluating density forecasts",
"authors": [
{
"first": "",
"middle": [],
"last": "Francis X Diebold",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "Todd",
"suffix": ""
},
{
"first": "Anthony S",
"middle": [],
"last": "Gunther",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Tay",
"suffix": ""
}
],
"year": 1997,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Francis X Diebold, Todd A Gunther, and Anthony S Tay. 1997. Evaluating density forecasts.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "Detecting visual text",
"authors": [
{
"first": "Jesse",
"middle": [],
"last": "Dodge",
"suffix": ""
},
{
"first": "Amit",
"middle": [],
"last": "Goyal",
"suffix": ""
},
{
"first": "Xufeng",
"middle": [],
"last": "Han",
"suffix": ""
},
{
"first": "Alyssa",
"middle": [],
"last": "Mensch",
"suffix": ""
},
{
"first": "Margaret",
"middle": [],
"last": "Mitchell",
"suffix": ""
},
{
"first": "Karl",
"middle": [],
"last": "Stratos",
"suffix": ""
},
{
"first": "Kota",
"middle": [],
"last": "Yamaguchi",
"suffix": ""
},
{
"first": "Yejin",
"middle": [],
"last": "Choi",
"suffix": ""
},
{
"first": "Hal",
"middle": [],
"last": "Daum\u00e9",
"suffix": ""
},
{
"first": "Iii",
"middle": [],
"last": "",
"suffix": ""
},
{
"first": "Alexander",
"middle": [
"C"
],
"last": "Berg",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of the NAACL-HLT",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Jesse Dodge, Amit Goyal, Xufeng Han, Alyssa Men- sch, Margaret Mitchell, Karl Stratos, Kota Yamaguchi, Yejin Choi, Hal Daum\u00e9 III, Alexander C Berg, et al. 2012. Detecting visual text. In Proceedings of the NAACL-HLT.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "Long-term recurrent convolutional networks for visual recognition and description",
"authors": [
{
"first": "Jeff",
"middle": [],
"last": "Donahue",
"suffix": ""
},
{
"first": "Lisa",
"middle": [
"Anne"
],
"last": "Hendricks",
"suffix": ""
},
{
"first": "Sergio",
"middle": [],
"last": "Guadarrama",
"suffix": ""
},
{
"first": "Marcus",
"middle": [],
"last": "Rohrbach",
"suffix": ""
},
{
"first": "Subhashini",
"middle": [],
"last": "Venugopalan",
"suffix": ""
},
{
"first": "Kate",
"middle": [],
"last": "Saenko",
"suffix": ""
},
{
"first": "Trevor",
"middle": [
"Darrell"
],
"last": "",
"suffix": ""
}
],
"year": 2014,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {
"arXiv": [
"arXiv:1411.4389"
]
},
"num": null,
"urls": [],
"raw_text": "Jeff Donahue, Lisa Anne Hendricks, Sergio Guadar- rama, Marcus Rohrbach, Subhashini Venugopalan, Kate Saenko, and Trevor Darrell. 2014. Long-term re- current convolutional networks for visual recognition and description. arXiv preprint arXiv:1411.4389.",
"links": null
},
"BIBREF13": {
"ref_id": "b13",
"title": "Copulas for information retrieval",
"authors": [
{
"first": "Carsten",
"middle": [],
"last": "Eickhoff",
"suffix": ""
},
{
"first": "P",
"middle": [],
"last": "Arjen",
"suffix": ""
},
{
"first": "Kevyn",
"middle": [],
"last": "De Vries",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Collins-Thompson",
"suffix": ""
}
],
"year": 2013,
"venue": "Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Carsten Eickhoff, Arjen P. de Vries, and Kevyn Collins- Thompson. 2013. Copulas for information retrieval. In Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Informa- tion Retrieval.",
"links": null
},
"BIBREF14": {
"ref_id": "b14",
"title": "Using the triangle inequality to accelerate k-means",
"authors": [
{
"first": "Charles",
"middle": [],
"last": "Elkan",
"suffix": ""
}
],
"year": 2003,
"venue": "ICML",
"volume": "3",
"issue": "",
"pages": "147--153",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Charles Elkan. 2003. Using the triangle inequality to accelerate k-means. In ICML, volume 3, pages 147- 153.",
"links": null
},
"BIBREF15": {
"ref_id": "b15",
"title": "From captions to visual concepts and back",
"authors": [
{
"first": "Saurabh",
"middle": [],
"last": "Hao Fang",
"suffix": ""
},
{
"first": "Forrest",
"middle": [],
"last": "Gupta",
"suffix": ""
},
{
"first": "Rupesh",
"middle": [],
"last": "Iandola",
"suffix": ""
},
{
"first": "Li",
"middle": [],
"last": "Srivastava",
"suffix": ""
},
{
"first": "Piotr",
"middle": [],
"last": "Deng",
"suffix": ""
},
{
"first": "Jianfeng",
"middle": [],
"last": "Doll\u00e1r",
"suffix": ""
},
{
"first": "Xiaodong",
"middle": [],
"last": "Gao",
"suffix": ""
},
{
"first": "Margaret",
"middle": [],
"last": "He",
"suffix": ""
},
{
"first": "John",
"middle": [],
"last": "Mitchell",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Platt",
"suffix": ""
}
],
"year": 2014,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {
"arXiv": [
"arXiv:1411.4952"
]
},
"num": null,
"urls": [],
"raw_text": "Hao Fang, Saurabh Gupta, Forrest Iandola, Rupesh Sri- vastava, Li Deng, Piotr Doll\u00e1r, Jianfeng Gao, Xi- aodong He, Margaret Mitchell, John Platt, et al. 2014. From captions to visual concepts and back. arXiv preprint arXiv:1411.4952.",
"links": null
},
"BIBREF16": {
"ref_id": "b16",
"title": "Every picture tells a story: Generating sentences from images",
"authors": [
{
"first": "Ali",
"middle": [],
"last": "Farhadi",
"suffix": ""
},
{
"first": "Mohsen",
"middle": [],
"last": "Hejrati",
"suffix": ""
},
{
"first": "Mohammad",
"middle": [
"Amin"
],
"last": "Sadeghi",
"suffix": ""
},
{
"first": "Peter",
"middle": [],
"last": "Young",
"suffix": ""
},
{
"first": "Cyrus",
"middle": [],
"last": "Rashtchian",
"suffix": ""
},
{
"first": "Julia",
"middle": [],
"last": "Hockenmaier",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Forsyth",
"suffix": ""
}
],
"year": 2010,
"venue": "Computer Vision-ECCV 2010",
"volume": "",
"issue": "",
"pages": "15--29",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ali Farhadi, Mohsen Hejrati, Mohammad Amin Sadeghi, Peter Young, Cyrus Rashtchian, Julia Hockenmaier, and David Forsyth. 2010. Every picture tells a story: Generating sentences from images. In Com- puter Vision-ECCV 2010, pages 15-29. Springer.",
"links": null
},
"BIBREF17": {
"ref_id": "b17",
"title": "Do we need hundreds of classifiers to solve real world classification problems",
"authors": [
{
"first": "Manuel",
"middle": [],
"last": "Fern\u00e1ndez-Delgado",
"suffix": ""
},
{
"first": "Eva",
"middle": [],
"last": "Cernadas",
"suffix": ""
},
{
"first": "Sen\u00e9n",
"middle": [],
"last": "Barro",
"suffix": ""
},
{
"first": "Dinani",
"middle": [],
"last": "Amorim",
"suffix": ""
}
],
"year": 2014,
"venue": "Journal of Machine Learning Research",
"volume": "15",
"issue": "",
"pages": "3133--3181",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Manuel Fern\u00e1ndez-Delgado, Eva Cernadas, Sen\u00e9n Barro, and Dinani Amorim. 2014. Do we need hun- dreds of classifiers to solve real world classification problems? Journal of Machine Learning Research, 15:3133-3181.",
"links": null
},
"BIBREF18": {
"ref_id": "b18",
"title": "Incorporating non-local information into information extraction systems by gibbs sampling",
"authors": [
{
"first": "Jenny",
"middle": [
"Rose"
],
"last": "Finkel",
"suffix": ""
},
{
"first": "Trond",
"middle": [],
"last": "Grenager",
"suffix": ""
},
{
"first": "Christopher",
"middle": [],
"last": "Manning",
"suffix": ""
}
],
"year": 2005,
"venue": "Proceedings of the 43rd Annual Meeting on Association for Computational Linguistics",
"volume": "",
"issue": "",
"pages": "363--370",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Jenny Rose Finkel, Trond Grenager, and Christopher Manning. 2005. Incorporating non-local information into information extraction systems by gibbs sampling. In Proceedings of the 43rd Annual Meeting on Associ- ation for Computational Linguistics, pages 363-370. Association for Computational Linguistics.",
"links": null
},
"BIBREF19": {
"ref_id": "b19",
"title": "Sampling-based approaches to calculating marginal densities",
"authors": [
{
"first": "Alan",
"middle": [],
"last": "Gelfand",
"suffix": ""
},
{
"first": "Adrian",
"middle": [],
"last": "Smith",
"suffix": ""
}
],
"year": 1990,
"venue": "Journal of the American statistical association",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Alan Gelfand and Adrian Smith. 1990. Sampling-based approaches to calculating marginal densities. Journal of the American statistical association.",
"links": null
},
"BIBREF20": {
"ref_id": "b20",
"title": "Copula-based kernel dependency measures",
"authors": [
{
"first": "Zoubin",
"middle": [],
"last": "Ghahramani",
"suffix": ""
},
{
"first": "Barnab\u00e1s",
"middle": [],
"last": "P\u00f3czos",
"suffix": ""
},
{
"first": "Jeff",
"middle": [],
"last": "Schneider",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of the 29th International Conference on Machine Learning",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Zoubin Ghahramani, Barnab\u00e1s P\u00f3czos, and Jeff Schnei- der. 2012. Copula-based kernel dependency mea- sures. In Proceedings of the 29th International Con- ference on Machine Learning.",
"links": null
},
"BIBREF21": {
"ref_id": "b21",
"title": "Understanding videos, constructing plots learning a visually grounded storyline model from annotated videos",
"authors": [
{
"first": "Abhinav",
"middle": [],
"last": "Gupta",
"suffix": ""
},
{
"first": "Praveen",
"middle": [],
"last": "Srinivasan",
"suffix": ""
},
{
"first": "Jianbo",
"middle": [],
"last": "Shi",
"suffix": ""
},
{
"first": "Larry S",
"middle": [],
"last": "Davis",
"suffix": ""
}
],
"year": 2009,
"venue": "Computer Vision and Pattern Recognition",
"volume": "",
"issue": "",
"pages": "2012--2019",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Abhinav Gupta, Praveen Srinivasan, Jianbo Shi, and Larry S Davis. 2009. Understanding videos, con- structing plots learning a visually grounded storyline model from annotated videos. In Computer Vision and Pattern Recognition, 2009. CVPR 2009. IEEE Confer- ence on, pages 2012-2019. IEEE.",
"links": null
},
"BIBREF22": {
"ref_id": "b22",
"title": "Coda: High dimensional copula discriminant analysis",
"authors": [
{
"first": "Fang",
"middle": [],
"last": "Han",
"suffix": ""
},
{
"first": "Tuo",
"middle": [],
"last": "Zhao",
"suffix": ""
},
{
"first": "Han",
"middle": [],
"last": "Liu",
"suffix": ""
}
],
"year": 2012,
"venue": "Journal of Machine Learning Research",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Fang Han, Tuo Zhao, and Han Liu. 2012. Coda: High dimensional copula discriminant analysis. Journal of Machine Learning Research.",
"links": null
},
"BIBREF23": {
"ref_id": "b23",
"title": "Improving neural networks by preventing coadaptation of feature detectors",
"authors": [
{
"first": "Nitish",
"middle": [],
"last": "Geoffrey E Hinton",
"suffix": ""
},
{
"first": "Alex",
"middle": [],
"last": "Srivastava",
"suffix": ""
},
{
"first": "Ilya",
"middle": [],
"last": "Krizhevsky",
"suffix": ""
},
{
"first": "Ruslan R",
"middle": [],
"last": "Sutskever",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Salakhutdinov",
"suffix": ""
}
],
"year": 2012,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {
"arXiv": [
"arXiv:1207.0580"
]
},
"num": null,
"urls": [],
"raw_text": "Geoffrey E Hinton, Nitish Srivastava, Alex Krizhevsky, Ilya Sutskever, and Ruslan R Salakhutdinov. 2012. Improving neural networks by preventing co- adaptation of feature detectors. arXiv preprint arXiv:1207.0580.",
"links": null
},
"BIBREF24": {
"ref_id": "b24",
"title": "Framing image description as a ranking task: Data, models and evaluation metrics",
"authors": [
{
"first": "Micah",
"middle": [],
"last": "Hodosh",
"suffix": ""
},
{
"first": "Peter",
"middle": [],
"last": "Young",
"suffix": ""
},
{
"first": "Julia",
"middle": [],
"last": "Hockenmaier",
"suffix": ""
}
],
"year": 2013,
"venue": "J. Artif. Intell. Res.(JAIR)",
"volume": "47",
"issue": "",
"pages": "853--899",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Micah Hodosh, Peter Young, and Julia Hockenmaier. 2013. Framing image description as a ranking task: Data, models and evaluation metrics. J. Artif. Intell. Res.(JAIR), 47:853-899.",
"links": null
},
"BIBREF25": {
"ref_id": "b25",
"title": "Introduction to mathematical statistics",
"authors": [
{
"first": "V",
"middle": [],
"last": "Robert",
"suffix": ""
},
{
"first": "Allen",
"middle": [],
"last": "Hogg",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Craig",
"suffix": ""
}
],
"year": 1994,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Robert V Hogg and Allen Craig. 1994. Introduction to mathematical statistics.",
"links": null
},
"BIBREF26": {
"ref_id": "b26",
"title": "Predicting popular messages in twitter",
"authors": [
{
"first": "Liangjie",
"middle": [],
"last": "Hong",
"suffix": ""
},
{
"first": "Ovidiu",
"middle": [],
"last": "Dan",
"suffix": ""
},
{
"first": "Brian",
"middle": [
"D"
],
"last": "Davison",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of WWW",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Liangjie Hong, Ovidiu Dan, and Brian D Davison. 2011. Predicting popular messages in twitter. In Proceedings of WWW.",
"links": null
},
"BIBREF27": {
"ref_id": "b27",
"title": "Multivariate models and dependence concepts",
"authors": [
{
"first": "Harry",
"middle": [],
"last": "Joe",
"suffix": ""
}
],
"year": 1997,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Harry Joe. 1997. Multivariate models and dependence concepts.",
"links": null
},
"BIBREF28": {
"ref_id": "b28",
"title": "Deep visualsemantic alignments for generating image descriptions",
"authors": [
{
"first": "Andrej",
"middle": [],
"last": "Karpathy",
"suffix": ""
},
{
"first": "Li",
"middle": [],
"last": "Fei-Fei",
"suffix": ""
}
],
"year": 2014,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Andrej Karpathy and Li Fei-Fei. 2014. Deep visual- semantic alignments for generating image descrip- tions. Stanford University Technical Report.",
"links": null
},
"BIBREF29": {
"ref_id": "b29",
"title": "A new measure of rank correlation",
"authors": [
{
"first": "Maurice",
"middle": [],
"last": "Kendall",
"suffix": ""
}
],
"year": 1938,
"venue": "Biometrika",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Maurice Kendall. 1938. A new measure of rank correla- tion. Biometrika.",
"links": null
},
"BIBREF30": {
"ref_id": "b30",
"title": "Imagenet classification with deep convolutional neural networks",
"authors": [
{
"first": "Alex",
"middle": [],
"last": "Krizhevsky",
"suffix": ""
},
{
"first": "Ilya",
"middle": [],
"last": "Sutskever",
"suffix": ""
},
{
"first": "Geoffrey",
"middle": [
"E"
],
"last": "Hinton",
"suffix": ""
}
],
"year": 2012,
"venue": "Advances in neural information processing systems",
"volume": "",
"issue": "",
"pages": "1097--1105",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. 2012. Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems, pages 1097-1105.",
"links": null
},
"BIBREF31": {
"ref_id": "b31",
"title": "Baby talk: Understanding and generating image descriptions",
"authors": [
{
"first": "Girish",
"middle": [],
"last": "Kulkarni",
"suffix": ""
},
{
"first": "Visruth",
"middle": [],
"last": "Premraj",
"suffix": ""
},
{
"first": "Sagnik",
"middle": [],
"last": "Dhar",
"suffix": ""
},
{
"first": "Siming",
"middle": [],
"last": "Li",
"suffix": ""
},
{
"first": "Yejin",
"middle": [],
"last": "Choi",
"suffix": ""
},
{
"first": "Alexander",
"middle": [
"C"
],
"last": "Berg",
"suffix": ""
},
{
"first": "Tamara",
"middle": [
"L"
],
"last": "Berg",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of the 24th CVPR",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Girish Kulkarni, Visruth Premraj, Sagnik Dhar, Siming Li, Yejin Choi, Alexander C Berg, and Tamara L Berg. 2011. Baby talk: Understanding and generating im- age descriptions. In Proceedings of the 24th CVPR. Citeseer.",
"links": null
},
"BIBREF32": {
"ref_id": "b32",
"title": "Automatic evaluation of information ordering: Kendall's tau. Computational Linguistics",
"authors": [
{
"first": "Mirella",
"middle": [],
"last": "Lapata",
"suffix": ""
}
],
"year": 2006,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mirella Lapata. 2006. Automatic evaluation of informa- tion ordering: Kendall's tau. Computational Linguis- tics.",
"links": null
},
"BIBREF33": {
"ref_id": "b33",
"title": "The nonparanormal: Semiparametric estimation of high dimensional undirected graphs",
"authors": [
{
"first": "Han",
"middle": [],
"last": "Liu",
"suffix": ""
},
{
"first": "John",
"middle": [],
"last": "Lafferty",
"suffix": ""
},
{
"first": "Larry",
"middle": [],
"last": "Wasserman",
"suffix": ""
}
],
"year": 2009,
"venue": "The Journal of Machine Learning Research",
"volume": "10",
"issue": "",
"pages": "2295--2328",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Han Liu, John Lafferty, and Larry Wasserman. 2009. The nonparanormal: Semiparametric estimation of high dimensional undirected graphs. The Journal of Machine Learning Research, 10:2295-2328.",
"links": null
},
"BIBREF34": {
"ref_id": "b34",
"title": "High-dimensional semiparametric gaussian copula graphical models. The Annals of Statistics",
"authors": [
{
"first": "Han",
"middle": [],
"last": "Liu",
"suffix": ""
},
{
"first": "Fang",
"middle": [],
"last": "Han",
"suffix": ""
},
{
"first": "Ming",
"middle": [],
"last": "Yuan",
"suffix": ""
},
{
"first": "John",
"middle": [],
"last": "Lafferty",
"suffix": ""
},
{
"first": "Larry",
"middle": [],
"last": "Wasserman",
"suffix": ""
}
],
"year": 2012,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Han Liu, Fang Han, Ming Yuan, John Lafferty, and Larry Wasserman. 2012. High-dimensional semiparamet- ric gaussian copula graphical models. The Annals of Statistics.",
"links": null
},
"BIBREF35": {
"ref_id": "b35",
"title": "Statistical language models based on neural networks",
"authors": [
{
"first": "Tom\u00e1\u0161",
"middle": [],
"last": "Mikolov",
"suffix": ""
}
],
"year": 2012,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Tom\u00e1\u0161 Mikolov. 2012. Statistical language models based on neural networks. Ph.D. thesis, Ph. D. the- sis, Brno University of Technology.",
"links": null
},
"BIBREF36": {
"ref_id": "b36",
"title": "Midge: Generating image descriptions from computer vision detections",
"authors": [
{
"first": "Margaret",
"middle": [],
"last": "Mitchell",
"suffix": ""
},
{
"first": "Xufeng",
"middle": [],
"last": "Han",
"suffix": ""
},
{
"first": "Jesse",
"middle": [],
"last": "Dodge",
"suffix": ""
},
{
"first": "Alyssa",
"middle": [],
"last": "Mensch",
"suffix": ""
},
{
"first": "Amit",
"middle": [],
"last": "Goyal",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of EACL",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Margaret Mitchell, Xufeng Han, Jesse Dodge, Alyssa Mensch, Amit Goyal, Alex Berg, Kota Yamaguchi, Tamara Berg, Karl Stratos, and Hal Daum\u00e9 III. 2012. Midge: Generating image descriptions from computer vision detections. In Proceedings of EACL.",
"links": null
},
"BIBREF37": {
"ref_id": "b37",
"title": "An introduction to copulas",
"authors": [
{
"first": "",
"middle": [],
"last": "Roger B Nelsen",
"suffix": ""
}
],
"year": 1999,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Roger B Nelsen. 1999. An introduction to copulas. Springer Verlag.",
"links": null
},
"BIBREF38": {
"ref_id": "b38",
"title": "Maltparser: A languageindependent system for data-driven dependency parsing",
"authors": [
{
"first": "Joakim",
"middle": [],
"last": "Nivre",
"suffix": ""
},
{
"first": "Johan",
"middle": [],
"last": "Hall",
"suffix": ""
},
{
"first": "Jens",
"middle": [],
"last": "Nilsson",
"suffix": ""
},
{
"first": "Atanas",
"middle": [],
"last": "Chanev",
"suffix": ""
},
{
"first": "G\u00fclsen",
"middle": [],
"last": "Eryigit",
"suffix": ""
},
{
"first": "Sandra",
"middle": [],
"last": "K\u00fcbler",
"suffix": ""
},
{
"first": "Svetoslav",
"middle": [],
"last": "Marinov",
"suffix": ""
},
{
"first": "Erwin",
"middle": [],
"last": "Marsi",
"suffix": ""
}
],
"year": 2007,
"venue": "Natural Language Engineering",
"volume": "13",
"issue": "02",
"pages": "95--135",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Joakim Nivre, Johan Hall, Jens Nilsson, Atanas Chanev, G\u00fclsen Eryigit, Sandra K\u00fcbler, Svetoslav Marinov, and Erwin Marsi. 2007. Maltparser: A language- independent system for data-driven dependency pars- ing. Natural Language Engineering, 13(02):95-135.",
"links": null
},
"BIBREF39": {
"ref_id": "b39",
"title": "Learning and transferring mid-level image representations using convolutional neural networks",
"authors": [
{
"first": "Maxime",
"middle": [],
"last": "Oquab",
"suffix": ""
},
{
"first": "Leon",
"middle": [],
"last": "Bottou",
"suffix": ""
},
{
"first": "Ivan",
"middle": [],
"last": "Laptev",
"suffix": ""
},
{
"first": "Josef",
"middle": [],
"last": "Sivic",
"suffix": ""
}
],
"year": 2013,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Maxime Oquab, Leon Bottou, Ivan Laptev, Josef Sivic, et al. 2013. Learning and transferring mid-level image representations using convolutional neural networks.",
"links": null
},
"BIBREF40": {
"ref_id": "b40",
"title": "Bleu: a method for automatic evaluation of machine translation",
"authors": [
{
"first": "Kishore",
"middle": [],
"last": "Papineni",
"suffix": ""
},
{
"first": "Salim",
"middle": [],
"last": "Roukos",
"suffix": ""
},
{
"first": "Todd",
"middle": [],
"last": "Ward",
"suffix": ""
},
{
"first": "Wei-Jing",
"middle": [],
"last": "Zhu",
"suffix": ""
}
],
"year": 2002,
"venue": "Proceedings of ACL",
"volume": "",
"issue": "",
"pages": "311--318",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kishore Papineni, Salim Roukos, Todd Ward, and Wei- Jing Zhu. 2002. Bleu: a method for automatic evalua- tion of machine translation. In Proceedings of ACL, pages 311-318. Association for Computational Lin- guistics.",
"links": null
},
"BIBREF41": {
"ref_id": "b41",
"title": "Copula regression. Variance Advancing and Science of Risk",
"authors": [
{
"first": "A",
"middle": [],
"last": "Rahul",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Parsa",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Stuart A Klugman",
"suffix": ""
}
],
"year": 2011,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Rahul A Parsa and Stuart A Klugman. 2011. Copula regression. Variance Advancing and Science of Risk.",
"links": null
},
"BIBREF42": {
"ref_id": "b42",
"title": "On estimation of a probability density function and mode. The annals of mathematical statistics",
"authors": [
{
"first": "Emanuel",
"middle": [],
"last": "Parzen",
"suffix": ""
}
],
"year": 1962,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Emanuel Parzen. 1962. On estimation of a probability density function and mode. The annals of mathemati- cal statistics.",
"links": null
},
"BIBREF43": {
"ref_id": "b43",
"title": "Efficient bayesian inference for gaussian copula regression models",
"authors": [
{
"first": "Michael",
"middle": [],
"last": "Pitt",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Chan",
"suffix": ""
},
{
"first": "Robert",
"middle": [],
"last": "Kohn",
"suffix": ""
}
],
"year": 2006,
"venue": "Biometrika",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Michael Pitt, David Chan, and Robert Kohn. 2006. Effi- cient bayesian inference for gaussian copula regression models. Biometrika.",
"links": null
},
"BIBREF44": {
"ref_id": "b44",
"title": "Probabilistic metric spaces",
"authors": [
{
"first": "Berthold",
"middle": [],
"last": "Schweizer",
"suffix": ""
},
{
"first": "Abe",
"middle": [],
"last": "Sklar",
"suffix": ""
}
],
"year": 1983,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Berthold Schweizer and Abe Sklar. 1983. Probabilistic metric spaces.",
"links": null
},
"BIBREF45": {
"ref_id": "b45",
"title": "Video google: A text retrieval approach to object matching in videos",
"authors": [
{
"first": "Josef",
"middle": [],
"last": "Sivic",
"suffix": ""
},
{
"first": "Andrew",
"middle": [],
"last": "Zisserman",
"suffix": ""
}
],
"year": 2003,
"venue": "Proceedings of ICCV",
"volume": "",
"issue": "",
"pages": "1470--1477",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Josef Sivic and Andrew Zisserman. 2003. Video google: A text retrieval approach to object matching in videos. In Proceedings of ICCV, pages 1470-1477. IEEE.",
"links": null
},
"BIBREF46": {
"ref_id": "b46",
"title": "Fonctions de r\u00e9partition\u00e0 n dimensions et leurs marges",
"authors": [
{
"first": "Abe",
"middle": [],
"last": "Sklar",
"suffix": ""
}
],
"year": 1959,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Abe Sklar. 1959. Fonctions de r\u00e9partition\u00e0 n dimen- sions et leurs marges. Universit\u00e9 Paris 8.",
"links": null
},
"BIBREF47": {
"ref_id": "b47",
"title": "Recursive deep models for semantic compositionality over a sentiment treebank",
"authors": [
{
"first": "Richard",
"middle": [],
"last": "Socher",
"suffix": ""
},
{
"first": "Alex",
"middle": [],
"last": "Perelygin",
"suffix": ""
},
{
"first": "Y",
"middle": [],
"last": "Jean",
"suffix": ""
},
{
"first": "Jason",
"middle": [],
"last": "Wu",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Chuang",
"suffix": ""
},
{
"first": "D",
"middle": [],
"last": "Christopher",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Manning",
"suffix": ""
},
{
"first": "Y",
"middle": [],
"last": "Andrew",
"suffix": ""
},
{
"first": "Christopher",
"middle": [],
"last": "Ng",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Potts",
"suffix": ""
}
],
"year": 2013,
"venue": "Proceedings of EMNLP",
"volume": "",
"issue": "",
"pages": "1631--1642",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Richard Socher, Alex Perelygin, Jean Y Wu, Jason Chuang, Christopher D Manning, Andrew Y Ng, and Christopher Potts. 2013. Recursive deep models for semantic compositionality over a sentiment treebank. In Proceedings of EMNLP, pages 1631-1642. Cite- seer.",
"links": null
},
"BIBREF48": {
"ref_id": "b48",
"title": "The effect of wording on message propagation: Topic-and author-controlled natural experiments on twitter",
"authors": [
{
"first": "Chenhao",
"middle": [],
"last": "Tan",
"suffix": ""
},
{
"first": "Lillian",
"middle": [],
"last": "Lee",
"suffix": ""
},
{
"first": "Bo",
"middle": [],
"last": "Pang",
"suffix": ""
}
],
"year": 2014,
"venue": "Proceedings of ACL",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Chenhao Tan, Lillian Lee, and Bo Pang. 2014. The ef- fect of wording on message propagation: Topic-and author-controlled natural experiments on twitter. In Proceedings of ACL.",
"links": null
},
"BIBREF49": {
"ref_id": "b49",
"title": "Regression shrinkage and selection via the lasso",
"authors": [
{
"first": "Robert",
"middle": [],
"last": "Tibshirani",
"suffix": ""
}
],
"year": 1996,
"venue": "Journal of the Royal Statistical Society. Series B (Methodological)",
"volume": "",
"issue": "",
"pages": "267--288",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Robert Tibshirani. 1996. Regression shrinkage and se- lection via the lasso. Journal of the Royal Statistical Society. Series B (Methodological), pages 267-288.",
"links": null
},
"BIBREF50": {
"ref_id": "b50",
"title": "Feature-rich part-of-speech tagging with a cyclic dependency network",
"authors": [
{
"first": "Kristina",
"middle": [],
"last": "Toutanova",
"suffix": ""
},
{
"first": "Dan",
"middle": [],
"last": "Klein",
"suffix": ""
},
{
"first": "D",
"middle": [],
"last": "Christopher",
"suffix": ""
},
{
"first": "Yoram",
"middle": [],
"last": "Manning",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Singer",
"suffix": ""
}
],
"year": 2003,
"venue": "Proceedings of NAACL-HLT",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kristina Toutanova, Dan Klein, Christopher D Manning, and Yoram Singer. 2003. Feature-rich part-of-speech tagging with a cyclic dependency network. In Pro- ceedings of NAACL-HLT.",
"links": null
},
"BIBREF51": {
"ref_id": "b51",
"title": "Show and tell: A neural image caption generator",
"authors": [
{
"first": "Oriol",
"middle": [],
"last": "Vinyals",
"suffix": ""
},
{
"first": "Alexander",
"middle": [],
"last": "Toshev",
"suffix": ""
},
{
"first": "Samy",
"middle": [],
"last": "Bengio",
"suffix": ""
},
{
"first": "Dumitru",
"middle": [],
"last": "Erhan",
"suffix": ""
}
],
"year": 2013,
"venue": "Advances in Neural Information Processing Systems",
"volume": "",
"issue": "",
"pages": "351--359",
"other_ids": {
"arXiv": [
"arXiv:1411.4555"
]
},
"num": null,
"urls": [],
"raw_text": "Oriol Vinyals, Alexander Toshev, Samy Bengio, and Du- mitru Erhan. 2014. Show and tell: A neural image caption generator. arXiv preprint arXiv:1411.4555. Stefan Wager, Sida Wang, and Percy Liang. 2013. Dropout training as adaptive regularization. In Ad- vances in Neural Information Processing Systems, pages 351-359.",
"links": null
},
"BIBREF52": {
"ref_id": "b52",
"title": "A semiparametric gaussian copula regression model for predicting financial risks from earnings calls",
"authors": [
{
"first": "Yang",
"middle": [],
"last": "William",
"suffix": ""
},
{
"first": "Zhenhao",
"middle": [],
"last": "Wang",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Hua",
"suffix": ""
}
],
"year": 2014,
"venue": "Proceedings of ACL",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "William Yang Wang and Zhenhao Hua. 2014. A semi- parametric gaussian copula regression model for pre- dicting financial risks from earnings calls. In Proceed- ings of ACL.",
"links": null
},
"BIBREF53": {
"ref_id": "b53",
"title": "Fast dropout training",
"authors": [
{
"first": "Sida",
"middle": [],
"last": "Wang",
"suffix": ""
},
{
"first": "Christopher",
"middle": [],
"last": "Manning",
"suffix": ""
}
],
"year": 2013,
"venue": "Proceedings of ICML",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Sida Wang and Christopher Manning. 2013. Fast dropout training. In Proceedings of ICML.",
"links": null
},
"BIBREF54": {
"ref_id": "b54",
"title": "Predicting a scientific community's response to an article",
"authors": [
{
"first": "Dani",
"middle": [],
"last": "Yogatama",
"suffix": ""
},
{
"first": "Michael",
"middle": [],
"last": "Heilman",
"suffix": ""
},
{
"first": "O'",
"middle": [],
"last": "Brendan",
"suffix": ""
},
{
"first": "Chris",
"middle": [],
"last": "Connor",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Dyer",
"suffix": ""
},
{
"first": "Noah A",
"middle": [],
"last": "Bryan R Routledge",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Smith",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of EMNLP",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Dani Yogatama, Michael Heilman, Brendan O'Connor, Chris Dyer, Bryan R Routledge, and Noah A Smith. 2011. Predicting a scientific community's response to an article. In Proceedings of EMNLP.",
"links": null
}
},
"ref_entries": {
"FIGREF0": {
"uris": null,
"type_str": "figure",
"text": "An example of the LOL cat memes.",
"num": null
},
"FIGREF1": {
"uris": null,
"type_str": "figure",
"text": "Our nonparanormal method extends Gaussian by transforming each dimension with a smooth function, and jointly models the stochastic dependencies among textual and visual features, as well as the popular votes by the crowd.",
"num": null
},
"FIGREF2": {
"uris": null,
"type_str": "figure",
"text": "Our pipeline for generating memes from raw images.",
"num": null
},
"FIGREF3": {
"uris": null,
"type_str": "figure",
"text": "Two figures on the left: varying the amount of training data. L(1): Spearman. L(2): Kendall. Two figures on the right: varying the amount of features. R(1): Spearman. R(2",
"num": null
},
"TABREF1": {
"num": null,
"content": "<table/>",
"type_str": "table",
"text": "The effects of dropout training for NPNs on meme and other datasets. The best results of each row are highlighted in bold. * indicates p < .001 comparing to the no dropout setting.",
"html": null
},
"TABREF3": {
"num": null,
"content": "<table/>",
"type_str": "table",
"text": "Top-30 linguistic features that are highly correlated with the popular votes.",
"html": null
},
"TABREF5": {
"num": null,
"content": "<table/>",
"type_str": "table",
"text": "The BLEU scores for generating memes from images. B-1 to B-4: BLEU unigram to four-grams. The best BLEU results are highlighted in bold. * indicates p < .001 comparing to the second best system.",
"html": null
}
}
}
} |