NoReC_sentence
This is a dataset for sentence-level sentiment classification in Norwegian, derived from the the fine-grained annotations of NoReC_fine. The original NoReC_fine dataset is described in the paper A Fine-Grained Sentiment Dataset for Norwegian by L. Øvrelid, P. Mæhlum, J. Barnes, and E. Velldal, published at LREC 2020. We here provide two new versions where the annotations have been aggregated at the sentence-level in the following ways:
- Binary: sentences that contained sentiment annotations of either positive or negative polarity (but not both).
- Three-way: additionaly includes sentences annotated as having no sentiment at all (neutral).
Note that for both versions, sentences that contained mixed polarity are excluded. The data comes with pre-defined train/dev/test splits, and the sentence counts for the various labels are as follows:
| Binary | Train | Dev | Test | Total |
|---|---|---|---|---|
| Positive | 2624 | 490 | 401 | 3515 |
| Negative | 1270 | 211 | 182 | 1663 |
| Three-way | Train | Dev | Test | Total |
|---|---|---|---|---|
| Positive | 2624 | 490 | 401 | 3515 |
| Negative | 1270 | 211 | 182 | 1663 |
| Neutral | 4079 | 710 | 598 | 5387 |
The data is released in JSON format, where every sentence is a dictionary with the following keys ('sent_id', 'text', 'label'):
{'sent_id': '201911-01-01', 'text': 'Philips 190G6', 'label': 'Neutral'}
The data can be easily loaded in the following way:
>>> import json
>>> data = {}
>>> for name in ["train", "dev", "test"]:
with open("binary/{0}.json".format(name)) as infile:
data[name] = json.load(infile)
Terms of use
NoReC_sentence inherits the license of the underlying NoReC corpus, copied here for convenience:
The data 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/
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.
Cite
If you use this dataset, please cite the following papers:
@InProceedings{KutBarVel21,
author = {Andrey Kutuzov and Jeremy Barnes and Erik Velldal and Lilja {\O}vrelid and Stephan Oepen},
title = {Large-Scale Contextualised Language Modelling for Norwegian},
booktitle = {{Proceedings of the 23rd Nordic Conference on Computational Linguistics (NoDaLiDa 2021)}},
year = 2021
}
@InProceedings{OvrMaeBar20,
author = {Lilja {\O}vrelid and Petter M{\ae}hlum and Jeremy Barnes and Erik Velldal},
title = {A Fine-grained Sentiment Dataset for {N}orwegian},
booktitle = {{Proceedings of the 12th Edition of the Language Resources and Evaluation Conference}},
year = 2020,
address = "Marseille, France, 2020"
}