egilron commited on
Commit
28187f0
1 Parent(s): 5ed3124

update readme

Browse files
Files changed (1) hide show
  1. README.md +46 -87
README.md CHANGED
@@ -69,30 +69,15 @@ configs:
69
  path: intensity/test-*
70
  ---
71
  # Dataset Card for NoReC TSA
 
72
 
 
73
 
74
- ## Dataset Description
75
-
76
- <!--- **Homepage:** --->
77
- - **Repository:**
78
- https://github.com/ltgoslo/norec_tsa
79
- - **Paper:**
80
- [A Fine-Grained Sentiment Dataset for Norwegian](https://aclanthology.org/2020.lrec-1.618/)
81
-
82
- <!---
83
- - **Leaderboard:**
84
- - **Point of Contact:**
85
- --->
86
-
87
- ### Dataset Summary
88
- The dataset contains tokenized Norwegian sentences where each token is tagged for sentiment expressed towards that token. The dataset is derived from the manually annotated [NoReC_fine](https://github.com/ltgoslo/norec_fine) with rich annotations for each sentiment expression in the texts.
89
- The texts are a subset of the Norewegian Review Corpus [NoReC](https://github.com/ltgoslo/norec).
90
-
91
- ### Supported Tasks and Leaderboards
92
- [NorBench](https://github.com/ltgoslo/norbench) provides TSA evaluation scripts using this dataset, and a leaderboard comparing large language models for downstream NLP tasks in Norwegian.
93
-
94
- ### Languages
95
- Norwegian: Predominantly Bokmål written variant.
96
 
97
  | variant | split | sents | docs |
98
  |:-----|:--------|--------:|-------:|
@@ -101,8 +86,14 @@ Norwegian: Predominantly Bokmål written variant.
101
  | nb | train | 8556 | 323 |
102
  | nn | train | 78 | 4 |
103
 
104
- ## Fine-tuned models
105
- [`ltg/norbert3-large_TSA`](https://huggingface.co/ltg/norbert3-large_TSA)
 
 
 
 
 
 
106
 
107
  ## Dataset Structure
108
  The dataset comes in two flavours:
@@ -127,16 +118,12 @@ Config "intensity" example instance:
127
  'tsa_tags': ['O', 'O', 'B-targ-Positive-2', 'O']}
128
  ```
129
 
130
-
131
-
132
  ### Data Fields
133
  - idx(str): Unique document-and sentence identifier from [NoReC_fine](https://github.com/ltgoslo/norec_fine). The 6-digit document identifier can also be used to look up the text and its metadata in [NoReC](https://github.com/ltgoslo/norec).
134
  - tokens: (List[str]): List of the tokens in the sentence
135
  - tsa_tags: (List[str]): List of the tags for each token in BIO format. There is no integer representation of these in the dataset.
136
 
137
 
138
-
139
-
140
  ### Data Splits
141
  ```
142
  DatasetDict({
@@ -155,80 +142,52 @@ DatasetDict({
155
  })
156
  ```
157
 
158
- ## Dataset Creation
159
-
160
- ### Curation Rationale
161
- The sentiment expressions and targets are annotated in NoReC_fine according to its [annotation guidelines](https://github.com/ltgoslo/norec_fine/blob/master/annotation_guidelines/guidelines.md)
162
-
163
- Since a sentiment target may be the target of several sentiment expressions, these are resolved to a final sentiment polarity (and intensity) using the conversion script in [NoReC_tsa](https://github.com/ltgoslo/norec_tsa). There is no "mixed" sentiment category. When a target is the receiver of both positive and negative sentiment, the strongest wins. If a tie, the last sentiment wins.
164
 
 
165
  ### Source Data
166
- A subset of the Norwegian Review Corpus with its sources and preprocessing described [here](https://github.com/ltgoslo/norec).
167
 
168
- <!---
169
- #### Initial Data Collection and Normalization
170
 
171
- [More Information Needed]
172
 
173
- #### Who are the source language producers?
174
 
175
- [More Information Needed]
176
 
177
- ### Annotations
178
 
179
- #### Annotation process
180
 
181
- [More Information Needed]
182
 
183
- #### Who are the annotators?
184
 
185
- [More Information Needed]
186
 
187
- ### Personal and Sensitive Information
188
-
189
- [More Information Needed]
190
-
191
- ## Considerations for Using the Data
192
-
193
- ### Social Impact of Dataset
194
-
195
- [More Information Needed]
196
- --->
197
- ### Discussion of Biases
198
- The professional review texts in NoReC that NoReC_tsa is a subset from, are from a set number of Norwegian Publishing channels and from a set timespan that can be explored in the [NoReC metadata](https://raw.githubusercontent.com/ltgoslo/norec/master/data/metadata.json). Both language usage and sentiments expressed could have been more diverse with a more diverse set of source texts.
199
-
200
- <!---
201
- ### Other Known Limitations
202
-
203
- [More Information Needed]
204
-
205
- ## Additional Information
206
-
207
- ### Dataset Curators
208
-
209
- [More Information Needed]
210
- --->
211
- ### Licensing Information
212
-
213
- The data, being derived from [NoReC](https://github.com/ltgoslo/norec), is distributed under a Creative Commons Attribution-NonCommercial licence (CC BY-NC 4.0), access the full license text here: https://creativecommons.org/licenses/by-nc/4.0/
214
 
215
- The licence is motivated by the need to block the possibility of third parties redistributing the orignal reviews for commercial purposes. Note that machine learned models, extracted lexicons, embeddings, and similar resources that are created on the basis of NoReC are not considered to contain the original data and so can be freely used also for commercial purposes despite the non-commercial condition.
 
 
 
 
 
 
 
 
216
 
 
217
 
218
- ### Citation Information
219
 
220
- ```bibtex
221
- @InProceedings{OvrMaeBar20,
222
- author = {Lilja Øvrelid and Petter Mæhlum and Jeremy Barnes and Erik Velldal},
223
- title = {A Fine-grained Sentiment Dataset for {N}orwegian},
224
- booktitle = {{Proceedings of the 12th Edition of the Language Resources and Evaluation Conference}},
225
- year = 2020,
226
- address = {Marseille, France, 2020}
227
- }
228
- ```
229
 
230
- <!---
231
- ### Contributions
232
 
233
- [More Information Needed]
234
- --->
 
69
  path: intensity/test-*
70
  ---
71
  # Dataset Card for NoReC TSA
72
+ This is a dataset for Targeted Sentiment Analysis (TSA) in Norwegian, derived from the the fine-grained annotations of N[NoReC_fine](https://github.com/ltgoslo/norec_fine). The dataset contains tokenized Norwegian sentences where each token is tagged for sentiment expressed within the sentence towards that token.
73
 
74
+ Since a sentiment target may be the target of several sentiment expressions, these are resolved to a final sentiment polarity (and intensity) using the conversion script in [NoReC_tsa](https://github.com/ltgoslo/norec_tsa). There is no "mixed" sentiment category. When a target is the receiver of both positive and negative sentiment, the strongest wins. If a tie, the last sentiment wins.
75
 
76
+ - **Curated by:** The [SANT](https://www.mn.uio.no/ifi/english/research/projects/sant/) project (Sentiment Analysis for Norwegian Text) at the [Language Technology Group](https://www.mn.uio.no/ifi/english/research/groups/ltg/) (LTG) at the University of Oslo
77
+ - **Funded by:** The [SANT](https://www.mn.uio.no/ifi/english/research/projects/sant/) is funded by the [Research Council of Norway](https://www.forskningsradet.no/en/) (NFR grant number 270908).
78
+ - **Shared by:** The [SANT](https://www.mn.uio.no/ifi/english/research/projects/sant/) project (Sentiment Analysis for Norwegian Text) at the [Language Technology Group](https://www.mn.uio.no/ifi/english/research/groups/ltg/) (LTG) at the University of Oslo
79
+ - **License:** The data is distributed under a [Creative Commons Attribution-NonCommercial licence](https://creativecommons.org/licenses/by-nc/4.0/) (CC BY-NC 4.0). The licence is motivated by the need to block the possibility of third parties redistributing the orignal reviews for commercial purposes. Note that machine learned models, extracted lexicons, embeddings, and similar resources that are created on the basis of NoReC are not considered to contain the original data and so can be freely used also for commercial purposes despite the non-commercial condition.
80
+ - **Language:** Norwegian ("no"): Predominantly Bokmål (nb) written variant.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
81
 
82
  | variant | split | sents | docs |
83
  |:-----|:--------|--------:|-------:|
 
86
  | nb | train | 8556 | 323 |
87
  | nn | train | 78 | 4 |
88
 
89
+ ## Dataset Sources
90
+
91
+ - **Repository:** https://github.com/ltgoslo/norec_tsa
92
+ - **Paper:** The underlying NoReC_fine dataset is described in the paper [A Fine-Grained Sentiment Dataset for Norwegian](https://aclanthology.org/2020.lrec-1.618/) by Øvrelid et al., published at LREC 2020.
93
+
94
+ ## Uses
95
+ The data is intended to be used for training and testing models for TSA token classification; identifying and classifying sentiment targets in Norwegian sentences.
96
+ Example models fine-tuned on this dataset can be found at [huggingface.co/collections/ltg/sentiment-analysis](https://huggingface.co/collections/ltg/sentiment-analysis-65c49c7247a0ffffa9897155)
97
 
98
  ## Dataset Structure
99
  The dataset comes in two flavours:
 
118
  'tsa_tags': ['O', 'O', 'B-targ-Positive-2', 'O']}
119
  ```
120
 
 
 
121
  ### Data Fields
122
  - idx(str): Unique document-and sentence identifier from [NoReC_fine](https://github.com/ltgoslo/norec_fine). The 6-digit document identifier can also be used to look up the text and its metadata in [NoReC](https://github.com/ltgoslo/norec).
123
  - tokens: (List[str]): List of the tokens in the sentence
124
  - tsa_tags: (List[str]): List of the tags for each token in BIO format. There is no integer representation of these in the dataset.
125
 
126
 
 
 
127
  ### Data Splits
128
  ```
129
  DatasetDict({
 
142
  })
143
  ```
144
 
 
 
 
 
 
 
145
 
146
+ ## Dataset Creation
147
  ### Source Data
 
148
 
149
+ The sentiment annotations are aggregated from the NoReC_fine dataset, which in turn comprises a subset of the documents in the [Norwegian Review Corpus](https://github.com/ltgoslo/norec) (NoReC), which contains full-text professional reviews collected from major Norwegian news sources and cover a range of different domains, including literature, movies, video games, restaurants, music and theater, in addition to product reviews across a range of categories. The review articles NoReC were originally donated by the media partners in the SANT project; the Norwegian Broadcasting Corporation (NRK), Schibsted Media Group and Aller Media. The data comprises reviews extracted from eight different Norwegian news sources: Dagbladet, VG, Aftenposten, Bergens Tidende, Fædrelandsvennen, Stavanger Aftenblad, DinSide.no and P3.no. In terms of publishing date the reviews of NoReC mainly cover the time span 2003–2019, although it also includes a handful of reviews dating back as far as 1998.
 
150
 
151
+ ### Annotators
152
 
153
+ The original annotations of NoReC_fine that the sentence-level labels here are derived from, were originally created by hired annotators who were all BSc- or MSc-level students in the Language Technology study program at the Department of informatics, University of Oslo.
154
 
155
+ #### Personal and Sensitive Information
156
 
157
+ <!-- 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. -->
158
 
159
+ The data does not contain information considered personal or sensitive.
160
 
161
+ ### Recommendations
162
 
163
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
164
 
165
+ Results obtained on this data might not generalize to texts from other domains or genres. Any biases in the sentiments expressed by the original review authors may carry over to models trained on this data.
166
 
167
+ ## Citation
168
+ **BibTeX:**
169
+ ```
170
+ @InProceedings{KutBarVel21,
171
+ author = {Andrey Kutuzov and Jeremy Barnes and Erik Velldal and Lilja {\O}vrelid and Stephan Oepen},
172
+ title = {Large-Scale Contextualised Language Modelling for Norwegian},
173
+ booktitle = {{Proceedings of the 23rd Nordic Conference on Computational Linguistics (NoDaLiDa 2021)}},
174
+ year = 2021
175
+ }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
176
 
177
+ @InProceedings{OvrMaeBar20,
178
+ author = {Lilja {\O}vrelid and Petter M{\ae}hlum and Jeremy Barnes and Erik Velldal},
179
+ title = {A Fine-grained Sentiment Dataset for {N}orwegian},
180
+ booktitle = {{Proceedings of the 12th Edition of the Language Resources and Evaluation Conference}},
181
+ year = 2020,
182
+ address = "Marseille, France, 2020"
183
+ }
184
+ ```
185
+ ## Dataset Card Authors
186
 
187
+ Egil Rønningstad and Erik Velldal
188
 
189
+ ## Dataset Card Contact
190
 
191
+ egilron@ifi.uio.no and erikve@ifi.uio.no
 
 
 
 
 
 
 
 
192
 
 
 
193