anilguven/bert_tr_turkish_product_reviews

Model Info

This model was developed/finetuned for product review task for Turkish Language. Model was finetuned via hepsiburada.com product review dataset.

• LABEL_0: negative review
• LABEL_1: positive review

Model Sources

• Dataset: https://huggingface.co/datasets/anilguven/turkish_product_reviews_sentiment
• Paper: https://ieeexplore.ieee.org/document/9559007
• Demo-Coding [optional]: https://github.com/anil1055/Turkish_Product_Review_Analysis_with_Language_Models
• Finetuned from model [optional]: https://huggingface.co/dbmdz/bert-base-turkish-cased

Preprocessing

You must apply removing stopwords, stemming, or lemmatization process for Turkish.

Results

• auprc = 0.9703364794020499
• auroc = 0.9740012964967856
• eval_loss = 0.358846469963511
• fn = 193
• fp = 207
• mcc = 0.8537512867685785
• tn = 2493
• tp = 2578
• Accuracy: %92.68

Citation

BibTeX:

@INPROCEEDINGS{9559007, author={Guven, Zekeriya Anil}, booktitle={2021 6th International Conference on Computer Science and Engineering (UBMK)}, title={The Effect of BERT, ELECTRA and ALBERT Language Models on Sentiment Analysis for Turkish Product Reviews}, year={2021}, volume={}, number={}, pages={629-632}, keywords={Computer science;Sentiment analysis;Analytical models;Computational modeling;Bit error rate;Time factors;Random forests;Sentiment Analysis;Language Model;Product Review;Machine Learning;E-commerce}, doi={10.1109/UBMK52708.2021.9559007}}

APA:

Guven, Z. A. (2021, September). The effect of bert, electra and albert language models on sentiment analysis for turkish product reviews. In 2021 6th International Conference on Computer Science and Engineering (UBMK) (pp. 629-632). IEEE.