davanstrien HF staff commited on
Commit
776b4e4
1 Parent(s): 37e2473

small fixes

Browse files
Files changed (1) hide show
  1. README.md +6 -6
README.md CHANGED
@@ -18,7 +18,7 @@ tags:
18
  - speeches
19
  - politics
20
  - parliament
21
- - british
22
  task_categories:
23
  - text-classification
24
  - text-generation
@@ -69,8 +69,8 @@ A dataset containing every speech in the House of Commons from May 1979-July 202
69
 
70
  ### Supported Tasks and Leaderboards
71
 
72
- - `text-classification`: This dataset can be used to clasify various texts (transcribed from speeches) as different time periods or as different types
73
- - `language-modeling`: This dataset can be used to contribute to the training or evaluation of language models for historical texts.
74
 
75
  ### Languages
76
 
@@ -142,13 +142,13 @@ Train: 2694375
142
 
143
  ### Curation Rationale
144
 
145
- This dataset contains all the speeches made in the House of Commons and can be used for a number of deep learning tasks like detecting how language and societal views have changed over the >40 years this dataset encompasses
146
 
147
  ### Source Data
148
 
149
  #### Initial Data Collection and Normalization
150
 
151
- The dataset is created by getting the data from [data.parliament.uk](http://data.parliament.uk/membersdataplatform/memberquery.aspx). There is no normalization
152
 
153
  #### Who are the source language producers?
154
 
@@ -176,7 +176,7 @@ The purpose of this dataset is to understand how language use and society's view
176
 
177
  ### Discussion of Biases
178
 
179
- Because of the long time period this dataset spans, it might contain language and opinions that are unacceptable in modern society
180
 
181
  ### Other Known Limitations
182
 
 
18
  - speeches
19
  - politics
20
  - parliament
21
+ - British
22
  task_categories:
23
  - text-classification
24
  - text-generation
 
69
 
70
  ### Supported Tasks and Leaderboards
71
 
72
+ - `text-classification`: This dataset can be used to classify various texts (transcribed from speeches) as different time periods or as different types
73
+ - `language-modeling`: This dataset can contribute to the training or the evaluation of language models for historical texts.
74
 
75
  ### Languages
76
 
 
142
 
143
  ### Curation Rationale
144
 
145
+ This dataset contains all the speeches made in the House of Commons and can be used for a number of deep learning tasks like detecting how language and societal views have changed over the >40 years. The dataset also provides language closer to the spoken language used in an elite British institution.
146
 
147
  ### Source Data
148
 
149
  #### Initial Data Collection and Normalization
150
 
151
+ The dataset is created by getting the data from [data.parliament.uk](http://data.parliament.uk/membersdataplatform/memberquery.aspx). There is no normalization.
152
 
153
  #### Who are the source language producers?
154
 
 
176
 
177
  ### Discussion of Biases
178
 
179
+ Because of the long time period this dataset spans, it might contain language and opinions that are unacceptable in modern society.
180
 
181
  ### Other Known Limitations
182