Fatima-Gh commited on
Commit
eebfcc8
β€’
1 Parent(s): cb7dccc

create readme

Browse files
Files changed (1) hide show
  1. README.md +159 -0
README.md ADDED
@@ -0,0 +1,159 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [![CC BY 4.0][cc-by-shield]][cc-by]
2
+
3
+
4
+ # GLARE: Google Apps Arabic Reviews
5
+
6
+
7
+
8
+ Dataset and Code of "GLARE: Google Apps Arabic Reviews" paper submitted to International Conference on Language Resources and Evaluation (LREC) 2022 to The 5th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT).
9
+
10
+
11
+ You can download the paper via: [[Github]](xx.pdf) [[DOI]](https://doi.org/xx/xx) [[ArXiv]](https://arxiv.org/abs/xxxx.xxxxx) [[PapersWithCode]](https://paperswithcode.com/).
12
+
13
+ ## Paper Summary
14
+
15
+ We introduce GLARE: Google Apps Arabic Reviews dataset. A collection of 76M reviews from 9,980 Android apps collected from Google PlayStore Saudi store.
16
+
17
+ ## Preparation
18
+ #### Below is details about each file, please ensure that you have enough storage before downloading the data.
19
+
20
+ | Data Type | File Name | File Size | File Type |
21
+ | ------------------ |---------------- | -------------- |-------------- |
22
+ | raw | apps | 4.1Β MB | CSV |
23
+ | raw | reviews | 25.3Β GB | CSV |
24
+ | raw | categories/ | 4.3 MB | CSV
25
+ | engineered | apps | 3.8Β MB | CSV
26
+ | engineered | reviews | 30.2 GB | CSV
27
+ | engineered | vocabulary | 530.5Β MB | CSV
28
+
29
+ ## File Specifications
30
+
31
+ - **apps.csv**: File that contains apps metadata.
32
+ - **reviews.csv**: File that contains reviews and reviews metadata.
33
+ - **categories/**: Folder that contains 59 CSV files, each file corresponds to one category with apps and apps metadata scrapped from top 200 free apps for that category.
34
+ - **vocabulary.csv**: File that contains vocabulary set generated from reviews with additional engineered features (word length, word frequency, has noise or digits, ..etc.)
35
+
36
+
37
+ ### Raw Data
38
+ #### Apps Metadata
39
+
40
+ ```
41
+ {
42
+ "title":"application name/title",
43
+ "app_id":"application unique identifier",
44
+ "url":"application url at Google PlayStore",
45
+ "icon":"url for image object",
46
+ "developer":"developer name",
47
+ "developer_id":"developer unique identifier",
48
+ "summary":"short description of the application",
49
+ "rating":"application accumulated rating"
50
+ }
51
+ ```
52
+
53
+ #### Reviews Metadata
54
+
55
+ ```
56
+
57
+ {
58
+ "at":"review datetime",
59
+ "content":"review text",
60
+ "replied_at":"developer reply datetime",
61
+ "reply_content":"developer reply content",
62
+ "review_created_version":"user application version during the time of review",
63
+ "review_id":"review unique identifier",
64
+ "rating":"user rating",
65
+ "thumbs_up_count":"number of users that agree with the reviewer",
66
+ "user_image":"url for image object",
67
+ "user_name":"user display name",
68
+ "app_id":"application unique identifier"
69
+ }
70
+
71
+
72
+ ```
73
+ ### Engineered Data
74
+
75
+ #### Apps Metadata
76
+ Same as apps.csv in raw data with the following additions:
77
+
78
+ ```
79
+ {
80
+ "reviews_count":"number of reviews for the application",
81
+ "categories":"list of application categories",
82
+ "categories_count":"number of application categories"
83
+
84
+ }
85
+ ```
86
+
87
+ #### Reviews Metadata
88
+ Same as reviews.csv in raw data with the following additions:
89
+
90
+ ```
91
+
92
+ {
93
+ "tokenized_review":"list of review words tokenized on white-space",
94
+ "words_count":"number of words in review"
95
+ }
96
+
97
+
98
+ ```
99
+
100
+
101
+ #### Vocabulary
102
+
103
+ ```
104
+
105
+ {
106
+ "word":"term text",
107
+ "length":"word characters count",
108
+ "frequency":"word occurrences in the reviews dataset",
109
+ "has_noise":"true or false if word contains anything non-arabic alphanumeric",
110
+ "noise":"list of noise (anything non-arabic alphanumeric) in the word",
111
+ "has_digits":"true or false if word contains arabic or hindi digits",
112
+ "digits":"list of digits in the word"
113
+ }
114
+
115
+
116
+ ```
117
+
118
+
119
+ ### Folders Structure
120
+
121
+ - Data are prepared as raw data or engineered data.
122
+ - Download the dataset files: [Google Drive](https://huggingface.co/datasets/Fatima-Gh/GLARE/) | [Zenodo](https://huggingface.co/datasets/Fatima-Gh/GLARE/)
123
+ - The directory structure is as follow:
124
+ ```
125
+ data
126
+ └── raw
127
+ β”œβ”€β”€ apps.csv
128
+ β”œβ”€β”€ reviews.csv
129
+ └── categories/
130
+ └── engineered
131
+ β”œβ”€β”€ apps.csv
132
+ β”œβ”€β”€ reviews.csv
133
+ └── vocabulary.csv
134
+ ```
135
+
136
+
137
+ ## Citation
138
+
139
+ If you use this dataset please cite as:
140
+
141
+ ```
142
+ @inproceedings{[author_first_name][year][abbr],
143
+ title={[paper title]},
144
+ author={[authors]},
145
+ booktitle={[venue]},
146
+ year={[year]}
147
+ }
148
+ ```
149
+
150
+ ## License
151
+
152
+ This work is licensed under a
153
+ [Creative Commons Attribution 4.0 International License][cc-by].
154
+
155
+ [![CC BY 4.0][cc-by-image]][cc-by]
156
+
157
+ [cc-by]: http://creativecommons.org/licenses/by/4.0/
158
+ [cc-by-image]: https://i.creativecommons.org/l/by/4.0/88x31.png
159
+ [cc-by-shield]: https://img.shields.io/badge/License-CC%20BY%204.0-lightgrey.svg