Research Article
Abstract
References
Information
기업들이 수집하고 있는 고객-상품 간 상관관계 데이터는 평점 데이터와 같은 Explicit Feedback 보다 상품 구매 여부처럼 선호도를 직접적으로 드러내지 않은 Implicit Feedback의 비중이 크다. 본 논문에서는 Implicit Feedback을 분석하기 위해 Explicit Feedback에서 사용하던 방식과 동일한 사용자 기반의 협업 필터링(User based Collaboration Filtering)을 적용한 결과, 고객 간 분포가 좁은 점으로 수렴하는 경향을 보여 추천 성능이 좋지 못한 것을 확인하였다. 데이터의 양이 상대적으로 많고 분산도가 높은 Implicit Feedback으로 상품 추천 성능과 연산 속도를 향상시키기 위해 ALS(Alternative Least Square)를 적용하여 국내 주요 유통사의 구매 데이터를 분석 및 학습한다. 학습한 결과, 고객 간 분포가 명확해진 것을 실험을 통해 확인하였다. 또한 학습된 모델을 이용하여 특정 고객에게 적합한 개인화 상품 추천 목록을 제시하고 검증 데이터로 실제 추천 목록을 구매한 이력이 있는지 확인한다. 검증을 통해 특정 고객의 분포가 많을수록 예측 정확도가 높은 것을 확인할 수 있었다.
The correlation data between customers and products collected by businesses consists of a significant portion of Implicit Feedback, such as product purchase history, which does not directly reveal preferences, unlike Explicit Feedback such as rating data. In this paper, when applying user-based collaborative filtering, which was previously used with Explicit Feedback, to analyze Implicit Feedback, it was observed that the recommendation performance suffered due to the convergence of customer distributions to narrow points. To improve the performance and computational speed of product recommendations using the relatively abundant and diverse Implicit Feedback data, ALS (Alternative Least Square) was applied to analyze and learn purchase data from major domestic retail companies. The results of the training process confirmed that the customer distributions became clearer. Additionally, the trained model was utilized to provide personalized product recommendation lists for specific customers, and validation data was used to verify whether the recommended lists were actually purchased. Through the validation, it was confirmed that higher prediction accuracy was achieved with customers who had more diverse distribution patterns.
- H.J. Yu, "추천 시스템", Communications of the Korean Institute of Information Scientists and Engineers, Vol. 32, No. 1, pp. 53-58, 2014.
- L. Greg, B. Smith, and J. York, "Amazon. com recommendations: Item-to-item collaborative filtering", IEEE Internet computing, Vol. 7, No. 1 pp. 76-80, 2003. 10.1109/MIC.2003.1167344
- 손지은, et al., "추천 시스템 기법 연구동향 분석", 대한산업공학회지, 제41권 제2호 pp. 185-208, 2015. 10.7232/JKIIE.2015.41.2.185
- X. He, et al., "Neural collaborative filtering", Proceedings of the 26th international conference on world wide web, 2017.10.1145/3038912.3052569
- J. Bennett, and L. Stan, "The netflix prize", Proceedings of KDD Cup and Workshop, Vol. 2007, 2007. 10.1145/1345448.1345459
- Y. Koren, R. Bell, and C. Volinsky, "Matrix factorization techniques for recommender systems", Computer, Vol. 42, No. 8, pp. 30-37, 2009. 10.1109/MC.2009.263
- M. Laurens, and G. Hinton, "Visualizing data using t-SNE", Journal of Machine Learning Research, Vol. 9, No. 11, 2008.
- Y. Hu, Y. Koren, and C. Volinsky, "Collaborative filtering for implicit feedback datasets", 2008 Eighth IEEE International Conference On Data Mining, IEEE, 2008. 10.1109/ICDM.2008.22 PMC2651197
- S. García, et al., "Big data preprocessing: methods and prospects", Big Data Analytics, Vol. 1, No. 1, pp. 1-22, 2016. 10.1186/s41044-016-0014-0
- S. Raschka, "Python machine learning", Packt Publishing LTD, 2015.
- Publisher :The Society of Convergence Knowledge
- Publisher(Ko) :융복합지식학회
- Journal Title :The Society of Convergence Knowledge Transactions
- Journal Title(Ko) :융복합지식학회논문지
- Volume : 11
- No :4
- Pages :1-10
- DOI :https://doi.org/10.22716/sckt.2023.11.4.031
The Society of Convergence Knowledge Transactions
