지능정보연구 (Journal of Intelligence and Information Systems)

  • 제29권4호
  • /
  • Pages.309-323
  • /
  • 2023
  • /
  • 2288-4866(pISSN)
  • /
  • 2288-4882(eISSN)
한국지능정보시스템학회 (Korea Intelligent Information System Society)
권호 표지
DOI QR코드

DOI QR Code

설명가능한 그래프 신경망을 활용한 리뷰 콘텐츠 기반의 유용성 예측모형

The Prediction of the Helpfulness of Online Review Based on Review Content Using an Explainable Graph Neural Network

  • 김은미 (부산대학교 경영연구원) ;
  • 야오즈옌 (부산대학교 경영학과BK21 디지털금융 교육연구단) ;
  • 홍태호 (부산대학교 경영대학)
  • Eunmi Kim (Institute of Management Research, Pusan National University) ;
  • Yao Ziyan (BK21 Digital Finance Education and Research Center, Pusan National University) ;
  • Taeho Hong (Business School, Pusan National University)
  • 투고 : 2023.12.06
  • 심사 : 2023.12.19
  • 발행 : 2023.12.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

온라인 리뷰의 역할이 중요해짐에 따라 유용한 리뷰를 선별하기 위해 많은 연구들이 이루어져 왔다. 유용한 리뷰는 고객들이 유용하다고 인지하는 리뷰이며, 평점, 리뷰길이, 리뷰내용 등에 영향을 받는 것으로 많은 연구에서 검증되었다. 유용한 리뷰는 소비자들의 투표에 의한 '좋아요' 수에 의해 결정되며 유용성 투표가 많을수록 소비자의 구매의사결정에 중요한 영향을 미치는 것으로 간주된다. 그러나 최근에 작성되어 많은 고객들에게 노출되지 않은 리뷰는 상대적으로 '좋아요' 수가 적을 수 있으며, 투표에 응하지 않아 '좋아요' 수가 없을 수도 있다. 따라서 유용한 리뷰를 판단하기 위해 '좋아요' 수에 의존하기 보다는 리뷰 내용을 기반으로 유용한 리뷰를 분류하고자 한다. 리뷰의 텍스트는 리뷰 유용성에 가장 큰 영향을 미치는 요인으로, 토픽 모델링, 감정분석 등 텍스트 마이닝 기법을 적용하여 리뷰 텍스트에 포함된 콘텐츠와 감정의 영향을 다양하게 분석하고 있다. 본 연구에서는 글로벌 영화정보 사이트인 IMDb의 영화리뷰를 활용하여 리뷰 콘텐츠 기반의 리뷰 유용성 예측모형을 제안한다. 설명가능한 그래프 신경망인 GNN(Graph Neural Network)을 적용하여 리뷰 유용성 예측모형을 구축하고, 설명가능한 인공지능을 통해 예측모형의 한계인 모형의 해석에 대한 문제를 해결한다. 설명가능한 그래프 신경망은 리뷰들 간의 연결관계도 확인할 수 있어 유용한 리뷰 또는 유용하지 않은 리뷰에 대해 보다 신뢰할 수 있는 정보를 제공할 수 있을 것이라 기대한다.

As the role of online reviews has become increasingly crucial, numerous studies have been conducted to utilize helpful reviews. Helpful reviews, perceived by customers, have been verified in various research studies to be influenced by factors such as ratings, review length, review content, and so on. The determination of a review's helpfulness is generally based on the number of 'helpful' votes from consumers, with more 'helpful' votes considered to have a more significant impact on consumers' purchasing decisions. However, recently written reviews that have not been exposed to many customers may have relatively few 'helpful' votes and may lack 'helpful' votes altogether due to a lack of participation. Therefore, rather than relying on the number of 'helpful' votes to assess the helpfulness of reviews, we aim to classify them based on review content. In addition, the text of the review emerges as the most influential factor in review helpfulness. This study employs text mining techniques, including topic modeling and sentiment analysis, to analyze the diverse impacts of content and emotions embedded in the review text. In this study, we propose a review helpfulness prediction model based on review content, utilizing movie reviews from IMDb, a global movie information site. We construct a review helpfulness prediction model by using an explainable Graph Neural Network (GNN), while addressing the interpretability limitations of the machine learning model. The explainable graph neural network is expected to provide more reliable information about helpful or non-helpful reviews as it can identify connections between reviews.

키워드

과제정보

이 논문은 2022년 대한민국 교육부와 한국연구재단의 지원을 받아 수행된 연구임 (NRF-2022S1A5B5A16055001)

참고문헌

  1. 김성수, 배준호, 이주현, 정희주, 김희웅. (2023). TeGCN: 씬파일러 신용평가를 위한 트랜스포머 임베딩 기반 그래프 신경망 구조 개발. 지능정보연구, 29(3), 419-437. https://doi.org/10.13088/JIIS.2023.29.3.419
  2. 류동엽, 이흠철, 김재경. (2023). XAI 기법을 이용한 리뷰 유용성 예측 결과 설명에 관한 연구. 지능정보연구, 29(2), 35-56. https://doi.org/10.13088/JIIS.2023.29.2.035
  3. 현우창, 이인수, 서봉원. (2023). 그래프 신경망을 활용한 온라인 의견 사기 탐지. 정보과학회논문지, 50(11), 985-994.
  4. 홍태호, 원종관, 김은미, 김민수. (2023). 설명 가능한 인공지능과 CNN을 활용한 암호화폐 가격 등락 예측모형. 지능정보연구, 29(2), 129-148. https://doi.org/10.13088/JIIS.2023.29.2.129
  5. Adak, A., Pradhan, B., & Shukla, N. (2022). Sentiment analysis of customer reviews of food delivery services using deep learning and explainable artificial intelligence: Systematic review. Foods, 11(10), 1500.
  6. Azzone, M., Barucci, E., Moncayo, G. G., & Marazzina, D. (2022). A machine learning model for lapse prediction in life insurance contracts. Expert Systems with Applications, 191, 116261.
  7. Cao, Q., Duan, W., & Gan, Q. (2011). Explorng determinants of voting for the "helpfulness" of online user reviews: A text mining approach. Decision Support Systems, 50, 511-521. https://doi.org/10.1016/j.dss.2010.11.009
  8. Coussement, K., & Benoit, D. F. (2021). Interpretable data science for decision making. Decision Support Systems, 150, 113664.
  9. Craja, P., Kim, A., & Lessmann, S. (2020). Deep learning for detecting financial statement fraud. Decision Support Systems, 139, 113421.
  10. Du, J., Rong, J., Wang, H., & Zhang, Y. (2021). Neighbor-aware review helpfulness prediction. Decision Support Systems, 148, 113581.
  11. Eslami, S. P., Ghasemaghaei, M., & Hassanein, K. (2018). Which online reviews do consumers find most helpful? A multi-method investigation. Decision Support Systems, 113, 32-42. https://doi.org/10.1016/j.dss.2018.06.012
  12. Fan, M., Feng, C., Guo, L., Sun, M., & Li, P. (2019, May). Product-aware helpfulness prediction of online reviews. In The world wide web conference (pp. 2715-2721).
  13. Fresneda, J. E., & Gefen, D. (2019). A semantic measure of online review helpfulness and the importance of message entropy. Decision Support Systems, 125, 113117.
  14. Ghose, A., & Ipeirotis, P. G. (2010). Estimating the helpfulness and economic impact of product reviews: Mining text and reviewer characteristics. IEEE transactions on knowledge and data engineering, 23(10), 1498-1512. https://doi.org/10.1109/TKDE.2010.188
  15. Hamilton, W., Ying, Z., & Leskovec, J. (2017). Inductive representation learning on large graphs. Advances in neural information processing systems, 30.
  16. Heng, Y., Gao, Z., Jiang, Y., & Chen, X. (2018). Exploring hidden factors behind online food shopping from Amazon reviews: A topic mining approach. Journal of Retailing and Consumer Services, 42, 161-168. https://doi.org/10.1016/j.jretconser.2018.02.006
  17. Hosaka, T. (2019). Bankruptcy prediction using imaged financial ratios and convolutional neural networks. Expert systems with applications, 117, 287-299. https://doi.org/10.1016/j.eswa.2018.09.039
  18. Jain, D. K., Rahate, A., Joshi, G., Walambe, R., & Kotecha, K. (2022). Employing Co-Learning to Evaluate the Explainability of Multimodal Sentiment Analysis. IEEE Transactions on Computational Social Systems.
  19. Jiang, W., & Luo, J. (2022). Graph neural network for traffic forecasting: A survey. Expert Systems with Applications, 207, 117921.
  20. Kipf, T. N., & Welling, M. (2016). Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907.
  21. Krishnamoorthy, S. (2015). Linguistic features for review helpfulness prediction. Expert Systems with Applications, 42(7), 3751-3759. https://doi.org/10.1016/j.eswa.2014.12.044
  22. Li, Z. (2022). Extracting spatial effects from machine learning model using local interpretation method: An example of SHAP and XGBoost. Computers, Environment and Urban Systems, 96, 101845.
  23. Liu, J., Chen, Y., Huang, X., Li, J., & Min, G. (2023). GNN-based long and short term preference modeling for next-location prediction. Information Sciences, 629, 1-14.
  24. Liu, Z., & Park, S. (2015). What makes a useful online review? Implication for travel product websites. Tourism management, 47, 140-151. https://doi.org/10.1016/j.tourman.2014.09.020
  25. Lundberg, S. M., Erion, G. G., & Lee, S. I. (2018). Consistent individualized feature attribution for tree ensembles. arXiv preprint arXiv:1802.03888.
  26. Mohammad, S. M. (2017). Word affect intensities. arXiv preprint arXiv:1704.08798.
  27. Mudambi, S. M., & Schuff, D. (2010). Research note: What makes a helpful online review? A study of customer reviews on Amazon. com. MIS quarterly, 34(1), 185-200. https://doi.org/10.2307/20721420
  28. Olmedilla, M., Martinez-Torres, M. R., & Toral, S. (2022). Prediction and modelling online reviews helpfulness using 1D Convolutional Neural Networks. Expert Systems with Applications, 198, 116787.
  29. Plutchik, R. (1980). Emotion: A Psychoevolutionary Synthesis. NY Harper and Row.
  30. Ribeiro, M. T., Singh, S., & Guestrin, C. (2016, August). "Why should i trust you?" Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, 1135-1144.
  31. Saptono, R., & Mine, T. (2022, October). Best Approximate Distribution-based Model for Helpful Vote of Customer Review Prediction. In 2022 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 3427-3434.
  32. Saumya, S., Roy, P. K., & Singh, J. P. (2023). Review helpfulness prediction on e-commerce websites: A comprehensive survey. Engineering Applications of Artificial Intelligence, 126, 107075.
  33. Saumya, S., Singh, J. P., Baabdullah, A. M., Rana, N. P., & Dwivedi, Y. K. (2018). Ranking online consumer reviews. Electronic commerce research and applications, 29, 78-89. https://doi.org/10.1016/j.elerap.2018.03.008
  34. Scarselli, F., Gori, M., Tsoi, A. C., Hagenbuchner, M., & Monfardini, G. (2008). The graph neural network model. IEEE transactions on neural networks, 20(1), 61-80.
  35. Shi, Z., Lee, G. M., & Whinston, A. B. (2016). Toward a Better Measure of Business Proximity. MIS quarterly, 40(4), 1035-1056. https://doi.org/10.25300/MISQ/2016/40.4.11
  36. Siering, M., Muntermann, J., & Rajagopalan, B. (2018). Explaining and predicting online review helpfulness: The role of content and reviewer-related signals. Decision Support Systems, 108, 1-12. https://doi.org/10.1016/j.dss.2018.01.004
  37. Wang, X., Tang, L. R., & Kim, E. (2019). More than words: Do emotional content and linguistic style matching matter on restaurant review helpfulness?. International Journal of Hospitality Management, 77, 438-447.
  38. Ying, Z., Bourgeois, D., You, J., Zitnik, M., & Leskovec, J. (2019). Gnnexplainer: Generating explanations for graph neural networks. Advances in neural information processing systems, 32.
  39. Zheng, T., Lin, Z., Zhang, Y., Jiao, Q., Su, T., Tan, H., ... & Law, R. (2023). Revisiting review helpfulness prediction: An advanced deep learning model with multimodal input from Yelp. International Journal of Hospitality Management, 114, 103579.
  40. Zhou, J., Cui, G., Hu, S., Zhang, Z., Yang, C., Liu, Z., ... & Sun, M. (2020). Graph neural networks: A review of methods and applications. AI open, 1, 57-81.