Utilizing quantification in sentiment analysis of product reviews
Abstract
The constant increase in online transactions has produced an enormous amount of information, such as reviews of products. These reviews gather the consumers’ sentiments about companies’ products, being strategic information for companies if correctly analyzed. Understanding the quantity of positive and negative reviews about products can be explored through quantification methods. In this paper, we evaluate different quantifiers applied to product reviews and assess the influence of these methods on classification performance. We compare ten quantification methods over six product review datasets. The results indicate that eight methods outperform the naı̈ve method for quantification tasks, and quantifiers can be employed to enhance the classification of product reviews. In both cases, statistically significant differences were observed.
References
Forman, G. (2005). Counting Positives Accurately Despite Inaccurate Classification. In Machine Learning: ECML 2005, volume 3720, pages 564–575. Springer Berlin Heidelberg, Berlin, Heidelberg.
Forman, G. (2007). Quantifying counts, costs, and trends accurately via machine learning. Technical report, Technical report, HP Laboratories, Palo Alto, CA.
González-Castro, V., Alaiz-Rodríguez, R., and Alegre, E. (2013). Class distribution estimation based on the Hellinger distance. Information Sciences, 218:146–164.
Griko Nibras (2018). Amazon Cell Phones Reviews. [link].
Hassan, W., Maletzke, A., and Batista, G. (2020). Accurately Quantifying a Billion Instances per Second. In 2020 IEEE 7th International Conference on Data Science and Advanced Analytics (DSAA), pages 1–10, Australia. IEEE.
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., and Stoyanov, V. (2019). RoBERTa: A Robustly Optimized BERT Pretraining Approach.
Maletzke, A., Dos Reis, D., Cherman, E., and Batista, G. (2019). DyS: A Framework for Mixture Models in Quantification. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01):4552–4560.
Maletzke, A., Hassan, W., dos Reis, D., and Batista, G. (2020). The Importance of the Test Set Size in Quantification Assessment. In Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, pages 2640–2646, Japan.
Medhat, W., Hassan, A., and Korashy, H. (2014). Sentiment analysis algorithms and applications: A survey. Ain Shams engineering journal, 5(4):1093–1113.
Ni, J., Li, J., and McAuley, J. (2019). Justifying Recommendations using Distantly-Labeled Reviews and Fine-Grained Aspects. In Proceedings of the Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 188–197, China.
Sebastiani, F. (2020). Evaluation measures for quantification: An axiomatic approach. Information Retrieval Journal, 23(3):255–288.
Tsytsarau, M. and Palpanas, T. (2012). Survey on mining subjective data on the web. Data Mining and Knowledge Discovery, 24:478–514.
