p1atdev commited on
Commit
2054e93
·
1 Parent(s): 59bd07b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +49 -27
README.md CHANGED
@@ -45,7 +45,6 @@ configs:
45
  - split: train
46
  path: livejupiter/train-*
47
  ---
48
-
49
  # おーぷん2ちゃんねる対話コーパス
50
 
51
  ## Dataset Details
@@ -64,44 +63,67 @@ configs:
64
 
65
  ## Dataset Structure
66
 
67
- <!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
68
-
69
- TODO
70
-
71
- [More Information Needed]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
72
 
73
  ## Dataset Creation
74
 
75
- ### Curation Rationale
76
-
77
- <!-- Motivation for the creation of this dataset. -->
78
-
79
- [More Information Needed]
80
-
81
  ### Source Data
82
 
83
- <!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
 
84
 
85
  #### Data Collection and Processing
86
 
87
- <!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
88
-
89
- [More Information Needed]
90
 
91
  #### Personal and Sensitive Information
92
 
93
- <!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
94
-
95
- [More Information Needed]
96
-
97
- ## Bias, Risks, and Limitations
98
-
99
- <!-- This section is meant to convey both technical and sociotechnical limitations. -->
100
 
101
- [More Information Needed]
102
 
103
- ### Recommendations
 
104
 
105
- <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
 
 
 
106
 
107
- Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
 
 
 
 
 
 
 
 
 
 
45
  - split: train
46
  path: livejupiter/train-*
47
  ---
 
48
  # おーぷん2ちゃんねる対話コーパス
49
 
50
  ## Dataset Details
 
63
 
64
  ## Dataset Structure
65
 
66
+ - `all-corpus`: `livejupiter`, `news4vip`, `newsplus` サブセットを連結したもの
67
+ - `dialogue`: 対話データ (`list[dict]`)
68
+ - `speaker`: 話者番号。`1` または `2`。
69
+ - `content`: 発言内容
70
+ - `board`: 連結元のサブセット名
71
+
72
+ - `livejupiter`: オリジナルのデータセットでの `livejupiter.tsv` から変換されたデータ。
73
+ - `dialogue`: 対話データ (`list[dict]`)
74
+ - `speaker`: 話者番号。`1` または `2`。
75
+ - `content`: 発言内容
76
+ - `news4vip`: オリジナルのデータセットでの `news4vip.tsv` から変換されたデータ。
77
+ - 構造は同上
78
+ - `newsplus`: オリジナルのデータセットでの `newsplus.tsv` から変換されたデータ。
79
+ - 構造は同上
80
+
81
+ - `ranking`: 応答順位付けタスク用データ (オリジナルデータセットでの `ranking.zip`)
82
+ - `train` と `test` split があり、それぞれはオリジナルデータセットの `dev.tsv` と `test.tsv` に対応します。
83
+ - `dialogue`: 対話データ (`list[dict]`)
84
+ - `speaker`: 話者番号。`1` または `2`。
85
+ - `content`: 発言内容
86
+ - `next`: 対話の次に続く正解の応答 (`dict`)
87
+ - `speaker`: 話者番号。`1` または `2`
88
+ - `content`: 発言内容
89
+ - `random`: ランダムに選ばれた応答 9 個 (`list[str]`)
90
+
91
+ また、`all-corpus`, `livejupiter`, `news4vip`, `newsplus` にはそれぞれ名前に `-cleaned` が付与されたバージョンがあり、これらのサブセットではオリジナルのデータセットで配布されていた NG ワードリストを利用してフィルタリングされたものです。
92
+
93
+ オリジナルのデータセットでは各発言内の改行は `__BR__` に置換されていますが、このデータセットではすべて `\n` に置き換えられています。
94
 
95
  ## Dataset Creation
96
 
 
 
 
 
 
 
97
  ### Source Data
98
 
99
+ (オリジナルのデータセットの説明より)
100
+ > おーぷん2ちゃんねるの「なんでも実況(ジュピター)」「ニュー速VIP」「ニュース速報+」の3つの掲示板をクロールして作成した対話コーパスです. おーぷん2ちゃんねる開設時から2019年7月20日までのデータを使用して作成しました.
101
 
102
  #### Data Collection and Processing
103
 
104
+ [オリジナルのデータセット](https://github.com/1never/open2ch-dialogue-corpus) を参照。
 
 
105
 
106
  #### Personal and Sensitive Information
107
 
108
+ `-cleaned` ではないサブセットでは、非常に不適切な表現が多いため注意が必要です。
 
 
 
 
 
 
109
 
110
+ ## Usage
111
 
112
+ ```py
113
+ from datasets import load_dataset
114
 
115
+ ds = load_dataset(
116
+ "p1atdev/open2ch",
117
+ name="all-corpus",
118
+ )
119
 
120
+ print(ds)
121
+ print(ds["train"][0])
122
+ # DatasetDict({
123
+ # train: Dataset({
124
+ # features: ['dialogue', 'board'],
125
+ # num_rows: 8134707
126
+ # })
127
+ # })
128
+ # {'dialogue': {'speaker': [1, 2], 'content': ['実況スレをたてる', 'おんj民の鑑']}, 'board': 'livejupiter'}
129
+ ```