Datasets:

Modalities:
Text
Formats:
text
Size:
< 1K
ArXiv:
Libraries:
Datasets
License:
zdy023 commited on
Commit
30150ff
·
1 Parent(s): 0adf04e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +54 -0
README.md CHANGED
@@ -1,3 +1,57 @@
1
  ---
2
  license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+
5
+ # WikiHow Task Set
6
+
7
+ The WikiHow task set is an InfoUI interaction task set based on
8
+ [Mobile-Env](https://github.com/X-LANCE/Mobile-Env) <!-- TODO: proposed in the
9
+ paper, the paper link -->. [WikiHow](https://www.wikihow.com/Main-Page) is a
10
+ collaborative wiki site about various real-life tips with more than 340,000
11
+ online articles. To construct the task set, 107,448 pages are crawled and the
12
+ dumped website data occupy about 88 GiB totally.
13
+
14
+ Several task definition templates are designed according to the functions of
15
+ the WikiHow app and 5,522 task definitions are instantiated through the
16
+ template toolkit in Mobile-Env. This task set is named as the *extended set*
17
+ (`wikihow-extended.tar.xz`). There may be several faults make the system or the
18
+ task fail in the auto-generated tasks. Therefore, 178 tasks are sampled from
19
+ the extended set and have been verified by human beings to ensure the
20
+ correctness and the stability, which is named as the *canonical set*
21
+ (`wikihow-canonical.tar.xz`). Owing to the limit of the budgets, only 70 tasks
22
+ are tested using the proposed LLM-based agent in the corresponding pager.
23
+ Thesed 70 tasks are given in `wikihow-microcanon.tar.xz`. We call it the
24
+ *canonical subset* or the *micro canonical set*.
25
+
26
+ ### Website Data Replay
27
+
28
+ The replay script for [mitmproxy](https://mitmproxy.org/) is given as
29
+ `replay_url.py`. To use this replay script, the information retrieval tool
30
+ [Pyserini](https://github.com/castorini/pyserini/) is in need. Four parameters
31
+ are expected to be assigned in the script:
32
+
33
+ + The crawled data from WikiHow website (`dumps` in `wikihow.data.tar.xz`)
34
+ + The html templates used for mocking the search result page (`templates` in
35
+ `wikihow.data.tar.xz`)
36
+ + The indices for the search engine based on Pyserini (`indices-t/indices` in
37
+ `wikihow.data.tar.xz`)
38
+ + The meta data of the crawled articles (`indices-t/docs/doc_meta.csv` in
39
+ `wikihow.data.tar.xz`)
40
+
41
+ All the data in need are offered in `wikihow.data.tar.xz`. (The archive is
42
+ about 78 GiB. And the decompressed data are about 88 GiB.)
43
+
44
+ To run the script:
45
+
46
+ ```sh
47
+ mitmproxy --showhost -s replay_url.py
48
+ ```
49
+
50
+ ### Certificate Unpinning Plan
51
+
52
+ The `syscert` plan proposed by Mobile-Env just works for WikiHow app. You can
53
+ complete the config according to the [guiding of
54
+ Mobile-Env](https://github.com/X-LANCE/Mobile-Env/blob/master/docs/dynamic-app-en.md).
55
+ The available apk package from [APKCombo](https://apkcombo.com/) is provided.
56
+ And note to used the AVD image of version Android 11.0 (API Level 30) (Google
57
+ APIs) to obtain the best compatibility and the root-enabled ADBD.