[논문]고객 선호 변화를 고려한 토픽 모델링 기반 추천 시스템

강소영; 김재경; 최일영; 강창동

doi:10.13088/jiis.2020.26.2.043

고객 선호 변화를 고려한 토픽 모델링 기반 추천 시스템
A Topic Modeling-based Recommender System Considering Changes in User Preferences 원문보기

지능정보연구 = Journal of intelligence and information systems, v.26 no.2, 2020년, pp.43 - 56

강소영 (경희대학교 대학원 경영학과) , 김재경 (경희대학교 경영학과) , 최일영 (경희대학교 경영대학원) , 강창동 (경희대학교 대학원 경영학과)

초록
AI-Helper

추천 시스템은 사용자가 다양한 옵션 중에서 최선의 선택을 할 수 있도록 도와준다. 그러나 추천 시스템이 상업적으로 성공하기 위해서는 극복할 몇 개의 문제점이 존재한다. 첫째, 추천시스템의 투명성 부족 문제이다. 즉, 추천된 상품이 왜 추천되었는지 사용자들이 알 수 없다. 둘째, 추천시스템이 사용자 선호의 변화를 즉각적으로 반영할 수 없는 문제이다. 즉, 사용자의 상품에 대한 선호는 시간이 지남에 따라 변함에도 불구하고, 추천시스템이 사용자 선호를 반영하기 위해서는 다시 모델을 재구축해야 한다. 따라서 본연구에서는 이러한 문제를 해결하기 위해 토픽 모델링과 순차 연관 규칙을 이용한 추천 방법론을 제안하였다. 토픽 모델링은 사용자에게 아이템이 왜 추천되었는지 설명하는데 유용하며, 순차 연관 규칙은 변화하는 사용자의 선호를 파악하는데 유용하다. 본 연구에서 제안한 방법은 크게 토픽 모델링 및 사용자 프로파일 생성 등 토픽 모델링에 기반한 사용자 프로파일 생성 단계와 토픽에 사용자 선호 확인 및 순차 연관 규칙 발견 등 순차 연관 규칙에 기반한 추천 단계로 구분된다. 벤치마크 시스템으로 협업 필터링 기반 추천 시스템을 개발하고, 아마존의 리뷰 데이터 셋을 이용하여 제안한 방법론의 성능을 비교 평가하였다. 비교 분석 결과, 제안한 방법론이 협업 필터링 기반 추천시스템보다 뛰어난 성능을 보였다. 따라서 본 연구에서 제안하는 추천 방법을 통해 추천 시스템의 투명성을 확보할 수 있을 뿐만 아니라, 시간에 따라 변화하는 사용자의 선호를 반영할 수 있다. 그러나 본 연구는 토픽과 관련된 상품을 추천하기 때문에, 토픽에 포함된 상품의 수가 많을 경우 추천이 정교하지 못하는 한계점이 있다. 또한 토픽의 수가 적기 때문에 토픽에 대한 순차 연관 규칙이 너무 적은 문제점이 있다. 향후 연구에서 이러한 문제점을 해결한다면 좋은 연구가 될 것으로 판단된다.

Abstract ▼ AI-Helper

Recommender systems help users make the best choice among various options. Especially, recommender systems play important roles in internet sites as digital information is generated innumerable every second. Many studies on recommender systems have focused on an accurate recommendation. However, there are some problems to overcome in order for the recommendation system to be commercially successful. First, there is a lack of transparency in the recommender system. That is, users cannot know why products are recommended. Second, the recommender system cannot immediately reflect changes in user preferences. That is, although the preference of the user's product changes over time, the recommender system must rebuild the model to reflect the user's preference. Therefore, in this study, we proposed a recommendation methodology using topic modeling and sequential association rule mining to solve these problems from review data. Product reviews provide useful information for recommendations because product reviews include not only rating of the product but also various contents such as user experiences and emotional state. So, reviews imply user preference for the product. So, topic modeling is useful for explaining why items are recommended to users. In addition, sequential association rule mining is useful for identifying changes in user preferences. The proposed methodology is largely divided into two phases. The first phase is to create user profile based on topic modeling. After extracting topics from user reviews on products, user profile on topics is created. The second phase is to recommend products using sequential rules that appear in buying behaviors of users as time passes. The buying behaviors are derived from a change in the topic of each user. A collaborative filtering-based recommendation system was developed as a benchmark system, and we compared the performance of the proposed methodology with that of the collaborative filtering-based recommendation system using Amazon's review dataset. As evaluation metrics, accuracy, recall, precision, and F1 were used. For topic modeling, collapsed Gibbs sampling was conducted. And we extracted 15 topics. Looking at the main topics, topic 1, top 3, topic 4, topic 7, topic 9, topic 13, topic 14 are related to "comedy shows", "high-teen drama series", "crime investigation drama", "horror theme", "British drama", "medical drama", "science fiction drama", respectively. As a result of comparative analysis, the proposed methodology outperformed the collaborative filtering-based recommendation system. From the results, we found that the time just prior to the recommendation was very important for inferring changes in user preference. Therefore, the proposed methodology not only can secure the transparency of the recommender system but also can reflect the user's preferences that change over time. However, the proposed methodology has some limitations. The proposed methodology cannot recommend product elaborately if the number of products included in the topic is large. In addition, the number of sequential patterns is small because the number of topics is too small. Therefore, future research needs to consider these limitations.

주제어

표/그림 (9)

그림 Overall procedure
표 Example of purchase history
표 Example of purchase history rearranged by topic
표 Example of user profile on topic
표 Example of behavior locus of user by time
그림 Data sample
그림 Distribution of the most frequently used words within each topic
표 Confusion matrix
그림 Comparative analysis of recommended performances

AI 본문요약
AI-Helper

* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.

가설 설정

However, the proposed methodology has some limitations. First, the proposed methodology does not recommend products directly but recommends products related to the topic. However, if there are too many products related to the topic, the sophistication of the recommendation may be deteriorated.

제안 방법

Amazon review data is used to evaluate the performance of the proposed methodology. The datasets are composed of the real-world reviews of 37,127 users who bought 1,685 amazon instant video from January 2007 to December 2013.
Our experiments were performed to compare the accuracy, recall, precision, and F1 of the proposed methodology with those of the benchmark system. The experimental result is presented in [Figure 3].
So, reviews imply user preference for the product. The proposed methodology employs topic modeling and sequential association rule mining. Topic modeling in product reviews is used for explainable recommendation because topic modeling estimates the distribution of topics per review (Pyo and Kim, 2014).

대상 데이터

Amazon review data is used to evaluate the performance of the proposed methodology. The datasets are composed of the real-world reviews of 37,127 users who bought 1,685 amazon instant video from January 2007 to December 2013. We deleted the data before 2010 because it was too old.

이론/모형

In this study, Latent Dirichlet Allocation (LDA) is used for topic modeling. To extract topics from user reviews, collapsed Gibbs sampling for LDA is used. Collapsed Gibbs sampling is a technique to remove some unnecessary variables from sampling.

성능/효과

Finally, the values of accuracy, recall, precision, and F1 on the benchmark system are 5.83%, 2.98%, 19.16%, 5.15%, respectively.
The experimental result is presented in [Figure 3]. First, the values of accuracy, recall, precision, and F1 on model 1 are 6.98%, 5.17%, 23.40%, and 8.47%, respectively. Second, the values of accuracy, recall, precision, and F1 on model 2 are 34.
47%, respectively. Second, the values of accuracy, recall, precision, and F1 on model 2 are 34.23%, 26.35%, 36.78%, and 30.71%, respectively. Finally, the values of accuracy, recall, precision, and F1 on the benchmark system are 5.
Therefore, we proposed a recommendation methodology using topic-modeling and sequential association rule mining. The experimental results show that the performance of the proposed methodology is better than that of the CF-based recommender system.

참고문헌 (35)

Abdollahi, B. and O. Nasraoui, "Explainable Matrix Factorization for Collaborative Filtering", In Proceedings of the 25th International Conference Companion on World Wide Web, (2016), 5-6.
Ahn, H. C., I. G. Han and K. J. Kim, "The Product Recommender System Combining Association Rules and Classification Models:The Case of G Internet Shopping Mall", Information Systems Review, Vol.8, No.1 (2006), 181-201.
Bellini, V., A. Schiavone, T. Di Noia, A. Ragone and E. Di Sciascio, "Knowledge-Aware Autoencoders for Explainable Recommender Systems", In Proceedings of the 3rd Workshop on Deep Learning for Recommender Systems, (2018), 24-31.
Bennett, J. and S. Lanning, "The Netflix Prize", In Proceedings of KDD cup and workshop, Vol.2007, (2007), 35.
Cho, Y. B., Y. H. Cho and S. H. Kim, "Mining Changes in Customer Buying Behavior for Collaborative Recommendations", Expert Systems with Applications, Vol.28, No.2 (2005), 359-369.

상세보기
Cho, Y. H., and J. K. Kim, "Application of Web Usage Mining and Product Taxonomy to Collaborative Recommendations in E-commerce", Expert systems with Applications, Vol.26, No.2 (2004), 233-246.

상세보기
Cho, Y. H., J. K. Kim and S. H. Kim, "A Personalized Recommender System based on Web Usage Mining and Decision Tree Induction", Expert systems with Applications, Vol.23, No.3 (2002), 329-342.

상세보기
Cho, Y. H., J. K. Kim, D. H. Ahn and H. A. Lee, "An Explanation based Recommender System using Collaborative Filtering: WebCF-Exp", Korean management Review, Vol.35, No.2 (2006), 493-519.
Choi, S., Y. Hyun and N. Kim, "Improving Performance of Recommendation Systems using Topic Modeling", Journal of Intelligence and Information Systems, Vol.21, No.3 (2015), 101-116.
Das, A. S., M. Datar, A. Garg and S. Rajaram, "Google News Personalization: Scalable Online Collaborative Filtering", In Proceedings of the 16th international conference on World Wide Web, (2007), 271-280.
Dehuri, S., Intelligent Techniques in Recommendation Systems: Contextual Advancements and New Methods: Contextual Advancements and New Methods, IGI Global, 2012.
Filieri, R. and F. McLeay, "E-WOM and Accommodation: An Analysis of the Factors That Influence Travelers' Adoption of Information from Online Reviews", Journal of Travel Research, Vol.53, No.1 (2014), 44-57.

상세보기
Good, N., J. B. Schafer, J. A. Konstan, A. Borchers, B. Sarwar, J. Herlocker and J. Riedl, "Combining Collaborative Filtering with Personal Agents for Better Recommendations", AAAI/IAAI, (1999), 439.
Gupta, P. and J. Harris, "How E-WOM Recommendations Influence Product Consideration and Quality of Choice: A Motivation to Process Information Perspective", Journal of Business Research, Vo.63, No.9-10 (2010), 1041-1049.
He, X., T. Chen, M. Y. Kan and X. Chen, "Trirank: Review-Aware Explainable Recommendation by Modeling Aspects", In Proceedings of the 24th ACM International on Conference on Information and Knowledge Management, (2015), 1661-1670.
Herlocker, J. L., J. A. Konstan and J. T. Riedl, "Explaining Collaborative Filtering Recommendations", In Proceedings of the 2000 ACM conference on Computer supported cooperative work, (2000), 241-250.
Herlocker, J. L., J. A. Konstan, L. G. Terveen and J. T. Riedl, "Evaluating Collaborative Filtering Recommender Systems", ACM Transactions on Information Systems (TOIS), Vol.22, No.1(2004), 5-53.
Blei, D. M. and J. D. Lafferty, "A Correlated Topic Model of Science", The Annals of Applied Statistics, Vol.1, No.1 (2007), 17-35.

상세보기
Jin, M., Luo, X., Zhu, H., and Zhuo, H. H., "Combining Deep Learning and Topic Modeling for Review Understanding in Context-aware Recommendation," In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Vol 1 (2018), 1605-1614.
Kim, H. K., J. K. Kim and Y. U. Ryu, "Personalized Recommendation over a Customer Network for Ubiquitous Shopping", IEEE Transactions on Services Computing, Vol.2, No.2 (2009), 140-151.

상세보기
Kim, H. N., A. T. Ji, I. Ha and G. S. Jo, "Collaborative Filtering based on Collaborative Tagging for Enhancing the Quality of Recommendation", Electronic Commerce Research and Applications, Vol.9, No.1 (2010), 73-83.

상세보기
Kim, J. K., H. S. Moon, B. J. An and I. Y. Choi, "A Grocery Recommendation for Off-line Shoppers", Online Information Review, Vol.42, No.4 (2018), 468-481.

상세보기
Lee, J. S. and S. D. Park, "Performance Improvement of a Movie Recommendation System using Genre-wise Collaborative Filtering", Journal of Intelligence and Information Systems, Vol.13, No.4 (2007), 65-78.
Moon, H. S., J. K. Kim and Y. U. Ryu, "A Sequence-based Filtering Method for Exhibition Booth Visit Recommendations", International journal of information management, Vol.33, No.4 (2013), 620-626.

상세보기
Moon, H. S., M. K. Jung, J. K. Kim and H. K. Kim, "Personalized Exhibition Booth Recommendation Methodology Using Sequential Association Rule", Journal of Intelligence and Information Systems, Vol.16, No.4 (2010), 195-211.
Ovsjanikov, M. and Y. Chen, "Topic Modeling for Personalized Recommendation of Volatile Items", In Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Springer, Berlin, Heidelberg, (2010), 483-498.
Pazzani, M. J. and D. Billsus, "Content-based Recommendation Systems", In The adaptive web, Springer, Berlin, Heidelberg, (2007), 325-341.
Ponweiser, M. Latent Dirichlet Allocation in R (Diploma thesis). Institute for Statistics and Mathematics, WU (Wirtschaftsuniversitat Wien), Austria, 2012.
Pyo, S. and E. Kim, "LDA-based Unified Topic Modeling for Similar TV User Grouping and TV Program Recommendation", IEEE transactions on cybernetics, Vol.45, No.8 (2014), 1476-1490.
Ren, Z., S. Liang, P. Li, S. Wang and M. de Rijke, "Social Collaborative Viewpoint Regression with Explainable Recommendations", In Proceedings of the tenth ACM international conference on web search and data mining, (2017), 485-494.
Sarwar, B., G. Karypis, J. A. Konstan and J. T. Riedl, "Analysis of Recommendation Algorithms for E-commerce", In Proceedings of the 2nd ACM conference on Electronic commerce, (2000), 158-167.
Shardanand, U. and P. Maes, "Social Information Filtering: Algorithms for Automating "Word of Mouth"", In Proceedings of the SIGCHI conference on Human factors in computing systems, (1995), 210-217.
Linden, G., B. Smith and J. York, "Amazon.com Recommendations: Item-to-Item Collaborative Filtering", IEEE Internet computing, Vol.7, No.1 (2003), 76-80.

상세보기
Sohn, J. B. and Y. M. Suh, "Using Degree of Match to Improve Prediction Quality in Collaborative Filtering Systems", Information Systems Review, Vol.8, No.2 (2006), 139-154.

원문보기 상세보기
Suh, K. S., S. Lee, E. K. Suh, H. Kang, S. Lee and U. K. Lee, "Comparisons of Popularityand Expert-Based News Recommendations:Similarities and Importance", Asia pacific journal of information systems, Vol.24, No.2 (2014), 191-210.

저자의 다른 논문 :

표제어: PCR

동의어: Packet Collision Rate

용어 설명 출처 목록 (6)

용어 설명: PCR은 세균 특이성이 있는 primer를 이용하여 적은 수의 세균이 있을지라도 쉽게 검출할 수 있는 유용한 방법이며, 이를 이용하여 구강 내 치면세균막이나 타액에서 직접 세균을 검출할 수 있게 되었다[8].

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증