File size: 13,555 Bytes
de28af9 |
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 |
import json
from pathlib import Path
from typing import Dict, List, Tuple
import datasets
from seacrowd.utils import schemas
from seacrowd.utils.configs import SEACrowdConfig
from seacrowd.utils.constants import Tasks, Licenses
_CITATION = """\
@article{kautsar2023indotod,
author={Kautsar, Muhammad Dehan Al and Nurdini, Rahmah Khoirussyifa' and Cahyawijaya, Samuel and Winata, Genta Indra and Purwarianti, Ayu},
title={IndoToD: A Multi-Domain Indonesian Benchmark For End-to-End Task-Oriented Dialogue Systems},
journal={arXiv preprint arXiv:2311.00958},
year={2023},
}
"""
_LANGUAGES = ["ind"]
_LOCAL = False
_DATASETNAME = "indosmd"
_DESCRIPTION = """\
IndoSMD is a synthetic task-oriented dialogue system dataset that was translated from the In-Car Assistant (SMD) dataset (Eric et al., 2017) into the new Indonesian dataset using the translation pipeline method
including delexicalization, translation, and delexicalization. The dataset consists of 323 dialogues in the POI Navigation, Calendar Scheduling, and Weather Information Retrieval domain, with a user and an agent talking to each other.
It also consists of slots and dialogue acts from the user and the agent.
"""
_HOMEPAGE = "https://github.com/dehanalkautsar/IndoToD/tree/main/IndoSMD"
_LICENSE = Licenses.CC_BY_SA_4_0.value
_URLS = {
_DATASETNAME: {
"train": "https://raw.githubusercontent.com/dehanalkautsar/IndoToD/main/IndoSMD/IndoSMD_split/IndoSMD_train.json",
"validation": "https://raw.githubusercontent.com/dehanalkautsar/IndoToD/main/IndoSMD/IndoSMD_split/IndoSMD_dev.json",
"test": "https://raw.githubusercontent.com/dehanalkautsar/IndoToD/main/IndoSMD/IndoSMD_split/IndoSMD_test.json",
},
}
_SUPPORTED_TASKS = [Tasks.E2E_TASK_ORIENTED_DIALOGUE]
_SOURCE_VERSION = "1.0.0"
_SEACROWD_VERSION = "2024.06.20"
class IndoSMDDataset(datasets.GeneratorBasedBuilder):
"""IndoToD: A Multi-Domain Indonesian Benchmark For End-to-End Task-Oriented Dialogue Systems"""
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION)
BUILDER_CONFIGS = [
SEACrowdConfig(
name=f"{_DATASETNAME}_source",
version=SOURCE_VERSION,
description="IndoToD: IndoSMD source schema",
schema="source",
subset_id=f"{_DATASETNAME}",
),
SEACrowdConfig(
name=f"{_DATASETNAME}_seacrowd_tod",
version=SEACROWD_VERSION,
description="IndoToD: IndoSMD SEACrowd End-to-end Task Oriented Dialogue schema",
schema="seacrowd_tod",
subset_id=f"{_DATASETNAME}",
),
]
DEFAULT_CONFIG_NAME = "indosmd_source"
def _info(self) -> datasets.DatasetInfo:
if self.config.schema == "source":
features = datasets.Features(
{
"index": datasets.Value("string"),
"dialogue": [
{
"turn": datasets.Value("string"),
"data": {
"end_dialogue": datasets.Value("string"),
"utterance": datasets.Value("string"),
"delex_utterance": datasets.Value("string"),
"requested": {
"distance": datasets.Value("string"),
"traffic_info": datasets.Value("string"),
"poi_type": datasets.Value("string"),
"address": datasets.Value("string"),
"poi": datasets.Value("string"),
"room": datasets.Value("string"),
"agenda": datasets.Value("string"),
"time": datasets.Value("string"),
"date": datasets.Value("string"),
"party": datasets.Value("string"),
"event": datasets.Value("string"),
"weather_attribute": datasets.Value("string"),
"location": datasets.Value("string"),
},
"slots": {
"distance": datasets.Value("string"),
"traffic_info": datasets.Value("string"),
"poi_type": datasets.Value("string"),
"address": datasets.Value("string"),
"poi": datasets.Value("string"),
"room": datasets.Value("string"),
"agenda": datasets.Value("string"),
"time": datasets.Value("string"),
"date": datasets.Value("string"),
"party": datasets.Value("string"),
"event": datasets.Value("string"),
"weather_attribute": datasets.Value("string"),
"location": datasets.Value("string"),
},
},
}
],
"scenario": {
"kb": {
"items": [
{
"distance": datasets.Value("string"),
"traffic_info": datasets.Value("string"),
"poi_type": datasets.Value("string"),
"address": datasets.Value("string"),
"poi": datasets.Value("string"),
"room": datasets.Value("string"),
"agenda": datasets.Value("string"),
"time": datasets.Value("string"),
"date": datasets.Value("string"),
"party": datasets.Value("string"),
"event": datasets.Value("string"),
"monday": datasets.Value("string"),
"tuesday": datasets.Value("string"),
"wednesday": datasets.Value("string"),
"thursday": datasets.Value("string"),
"friday": datasets.Value("string"),
"saturday": datasets.Value("string"),
"sunday": datasets.Value("string"),
"today": datasets.Value("string"),
"location": datasets.Value("string"),
}
],
"column_names": [datasets.Value("string")],
"kb_title": datasets.Value("string"),
},
"task": {"intent": datasets.Value("string")},
"uuid": datasets.Value("string"),
},
}
)
elif self.config.schema == "seacrowd_tod":
features = schemas.tod_features
else:
raise NotImplementedError(f"Schema {self.config.schema} has not been implemented")
return datasets.DatasetInfo(
description=_DESCRIPTION,
features=features,
homepage=_HOMEPAGE,
license=_LICENSE,
citation=_CITATION,
)
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
"""Returns SplitGenerators."""
urls = _URLS[_DATASETNAME]
data_dir = dl_manager.download_and_extract(urls)
return [
datasets.SplitGenerator(
name=datasets.Split.TRAIN,
gen_kwargs={
"filepath": data_dir["train"],
"split": "train",
},
),
datasets.SplitGenerator(
name=datasets.Split.VALIDATION,
gen_kwargs={
"filepath": data_dir["validation"],
"split": "validation",
},
),
datasets.SplitGenerator(
name=datasets.Split.TEST,
gen_kwargs={
"filepath": data_dir["test"],
"split": "test",
},
),
]
def _generate_examples(self, filepath: Path, split: str) -> Tuple[int, Dict]:
"""Yields examples as (key, example) tuples."""
key_slot_constant = ["distance", "traffic_info", "poi_type", "address", "poi", "room", "agenda", "time", "date", "party", "event", "weather_attribute", "location"]
key_kb_constant = ["distance", "traffic_info", "poi_type", "address", "poi", "room", "agenda", "time", "date", "party", "event", "monday", "tuesday", "wednesday", "thursday", "friday", "saturday", "sunday", "today", "location"]
with open(filepath, "r+") as fw:
data = json.loads(fw.read())
if self.config.schema == "source":
for idx, example in enumerate(data):
example["index"] = str(idx)
for i in range(len(example["dialogue"])):
if "requested" not in example["dialogue"][i]["data"]: # the difference between user and system utterance (user and system utterance is divided into each dict in the origin dataset)
example["dialogue"][i]["data"]["requested"] = {}
example["dialogue"][i]["data"]["slots"] = {}
for key in key_slot_constant:
example["dialogue"][i]["data"]["requested"][key] = ""
example["dialogue"][i]["data"]["slots"][key] = ""
else:
for key in key_slot_constant:
if key not in example["dialogue"][i]["data"]["requested"]:
example["dialogue"][i]["data"]["requested"][key] = ""
if key not in example["dialogue"][i]["data"]["slots"]:
example["dialogue"][i]["data"]["slots"][key] = ""
if not example["scenario"]["kb"].get("items"):
example["scenario"]["kb"]["items"] = []
for i in range(len(example["scenario"]["kb"]["items"])):
for key in key_kb_constant:
if key not in example["scenario"]["kb"]["items"][i]:
example["scenario"]["kb"]["items"][i][key] = ""
yield str(idx), example
elif self.config.schema == "seacrowd_tod":
for idx, tod_dialogue in enumerate(data):
example = {}
example["dialogue_idx"] = idx
dialogue = []
# NOTE: the dialogue always started with `driver` as first utterance
for turn, i in enumerate(range(0, len(tod_dialogue["dialogue"]) + 2, 2)):
dial = {}
dial["turn_idx"] = turn
# system_utterance properties
dial["system_utterance"] = ""
dial["system_acts"] = []
if turn != 0:
dial["system_utterance"] = tod_dialogue["dialogue"][i - 1]["data"]["utterance"]
if i < len(tod_dialogue["dialogue"]):
# NOTE: system_acts will be filled with every slot that has 'True' value on the origin dataset (on the requested field)
for act in tod_dialogue["dialogue"][i + 1]["data"]["requested"]:
if tod_dialogue["dialogue"][i + 1]["data"]["requested"][act]:
dial["system_acts"].append([act])
# user_utterance properties
dial["turn_label"] = []
dial["belief_state"] = []
if i == len(tod_dialogue["dialogue"]):
# case if turn_idx > len(dialogue) --> add dummy user_utterance
dial["user_utterance"] = ""
else:
dial["user_utterance"] = tod_dialogue["dialogue"][i]["data"]["utterance"]
# NOTE: belief_state will be filled with request act from `requested` field & inform act from `slots` field in the origin dataset
for act in tod_dialogue["dialogue"][i + 1]["data"]["requested"]:
if tod_dialogue["dialogue"][i + 1]["data"]["requested"][act]:
dial["belief_state"].append({"slots": [["slot", act]], "act": "request"})
for slot, slot_value in tod_dialogue["dialogue"][i + 1]["data"]["slots"].items():
dial["belief_state"].append({"slots": [[slot, slot_value]], "act": "inform"})
# append to dialogue
dialogue.append(dial)
example["dialogue"] = dialogue
yield str(idx), example
|