상품 및 서비스에 대한 개인화된 추천 서비스가 중요해짐에 따라, 많은 연구자들은 추천시스템 개발을 위한 다양한 지식과 기법들을 제공해왔다. 이러한 기법들 중에서 협업 필터링(Collaborative Fitering) 기법은 여러 분야에서 널리 사용되고 있으며, 그 유용성이 입증되었다. 하지만, 추천시스템의 성능을 더욱 높이기 위해서 현재의 협업 필터링 기법은 다음과 같은 점들을 고려해야 한다. 첫째, 대부분의 추천시스템과 관련한 연구에서 특정 고객과 성향이 유사한 다른 고객들을 찾기 위해 사용되는 유사도 함수들(Similarity Functions)은 대부분 특정한 관점에 초점을 두고 있기 때문에 다양한 관점에서 성향이 유사한 다른 고객들을 찾는데 한계를 가진다. 따라서, 특정 관점에 치우치지 않는 통합된 유사도 함수를 사용해야 할 필요가 있다. 둘째, 고객들의 성향은 시간이 지남에 따라 변화하기 때문에, 이를 추천결과에 반영하기 위해서는 시간에 따른 고객들의 구매 성향의 변화를 고려해야 한다. 본 연구는 여러 실험들을 통해 다음의 가설을 검정하는 것을 목적으로 하였다-다양한 관점이 동시에 반영된 통합 유사도 함수의 이용과 시간정보를 이용한 사용자의 구매 성향의 변화를 반영할 경우 추천의 정확도가 향상될 것이다. 다양한 실험을 통해, 본 연구에서 제시한 추천시스템은 전통적인 협업 필터링 기반의 추천시스템들에 비해 일반적으로 상당히 높은 정확도를 보였으며 이를 통해, 본 연구에서 제시한 가설이 채택될 수 있음을 확인하였다.
상품 및 서비스에 대한 개인화된 추천 서비스가 중요해짐에 따라, 많은 연구자들은 추천시스템 개발을 위한 다양한 지식과 기법들을 제공해왔다. 이러한 기법들 중에서 협업 필터링(Collaborative Fitering) 기법은 여러 분야에서 널리 사용되고 있으며, 그 유용성이 입증되었다. 하지만, 추천시스템의 성능을 더욱 높이기 위해서 현재의 협업 필터링 기법은 다음과 같은 점들을 고려해야 한다. 첫째, 대부분의 추천시스템과 관련한 연구에서 특정 고객과 성향이 유사한 다른 고객들을 찾기 위해 사용되는 유사도 함수들(Similarity Functions)은 대부분 특정한 관점에 초점을 두고 있기 때문에 다양한 관점에서 성향이 유사한 다른 고객들을 찾는데 한계를 가진다. 따라서, 특정 관점에 치우치지 않는 통합된 유사도 함수를 사용해야 할 필요가 있다. 둘째, 고객들의 성향은 시간이 지남에 따라 변화하기 때문에, 이를 추천결과에 반영하기 위해서는 시간에 따른 고객들의 구매 성향의 변화를 고려해야 한다. 본 연구는 여러 실험들을 통해 다음의 가설을 검정하는 것을 목적으로 하였다-다양한 관점이 동시에 반영된 통합 유사도 함수의 이용과 시간정보를 이용한 사용자의 구매 성향의 변화를 반영할 경우 추천의 정확도가 향상될 것이다. 다양한 실험을 통해, 본 연구에서 제시한 추천시스템은 전통적인 협업 필터링 기반의 추천시스템들에 비해 일반적으로 상당히 높은 정확도를 보였으며 이를 통해, 본 연구에서 제시한 가설이 채택될 수 있음을 확인하였다.
As personalized recommendation of products and services is rapidly growing in importance, a number of studies provided fundamental knowledge and techniques for developing recommendation systems. Among them, the CF technique has been most widely used and has proven to be useful in many practices. How...
As personalized recommendation of products and services is rapidly growing in importance, a number of studies provided fundamental knowledge and techniques for developing recommendation systems. Among them, the CF technique has been most widely used and has proven to be useful in many practices. However, current collaborative filtering (CF) technique has still considerable rooms for improving the effectiveness of recommendation systems: 1) a similarity function most systems use to find so-called like-minded people is not well defined in that similarity is computed from a single perspective of similarity concept; and 2) temporal information that contains the changing preference of customers needs to be taken into account when making recommendations. We hypothesize that integration of multiple aspects of similarity and utilization of temporal information will improve the accuracy of recommendations. The objective of this paper is to test the hypothesis through a series of experiments using MovieLens data. The experimental results show that the proposed recommendation system highly outperforms the conventional CF-based systems, confirming our hypothesis.
As personalized recommendation of products and services is rapidly growing in importance, a number of studies provided fundamental knowledge and techniques for developing recommendation systems. Among them, the CF technique has been most widely used and has proven to be useful in many practices. However, current collaborative filtering (CF) technique has still considerable rooms for improving the effectiveness of recommendation systems: 1) a similarity function most systems use to find so-called like-minded people is not well defined in that similarity is computed from a single perspective of similarity concept; and 2) temporal information that contains the changing preference of customers needs to be taken into account when making recommendations. We hypothesize that integration of multiple aspects of similarity and utilization of temporal information will improve the accuracy of recommendations. The objective of this paper is to test the hypothesis through a series of experiments using MovieLens data. The experimental results show that the proposed recommendation system highly outperforms the conventional CF-based systems, confirming our hypothesis.
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
문제 정의
Based on these ideas, we hypothesize that when multiple similarities computed from different perspectives are integrated and temporal information is utilized, the resulting recommendation quality and value will be improved. Therefore, the objective of this paper is to test the hypothesis through experiments using MovieLens data. We first located items to recommend for individual users, considering three aspects of similarity when we sought for like‐minded neighbors and then checked whether or not each user had watched the recommended movies by a specific point of time after the recommendation was made.
This experiment confirmed our idea that users’ tastes and preferences change over time and users have more interest in newly purchased items and/or newly launched items.
However, they still have considerable rooms for improvement in terms of recommendation quality and value. This paper proposes a novel approach to improving them. First, we suggest a new similarity function which reflects multiple aspects of similarity concept.
가설 설정
Based on these ideas, we hypothesize that when multiple similarities computed from different perspectives are integrated and temporal information is utilized, the resulting recommendation quality and value will be improved. Therefore, the objective of this paper is to test the hypothesis through experiments using MovieLens data.
As for the second idea of integrating temporal information, we used two kinds of temporal information, purchasing time and item launch time, in the recommendation process of CF technique. We hypothesize that when multiple similarity perspectives are integrated and temporal information is utilized, the resulting recommendation quality will be improved.
제안 방법
Therefore, to lessen the absurdity resulting from using a single similarity function, we integrated the three similarity scores computed from three different perspectives (i.e., pattern, angle and magnitude) into one. Another thing that we took into account while integrating them is that a user’s rating information may be more reliable if the user rated more items that a target user rated than other users did.
This research proposed a new novel approach to improving the accuracy of current CF‐based recommendation system, based on two ideas of integrating different similarity perspectives and of integrating temporal information.
대상 데이터
, the last 16 months) were used for testing. 20 target users were randomly selected based on whether they had watched movies more than 20 times in the training data, and had watched a movie within the first six months in the test data. For the evaluation of our recommendation system, we compared the movies the target users had watched during the entire test period with the movies our system recommended.
성능/효과
2) each item should be highly rated by users whose profiles are similar to the target user’ profile.
참고문헌 (51)
Adomavicius, G., "Towards the Next Generation of Recommender Systems: A Survey of the StateoftheArt and Possible Extensions", IEEE Transactions on Knowledge and Data Engineering, Vol.17, No.6(2005), 734-749.
Adomavicius, G. and Y. Kwon, "New Recommendation Techniques for Multicriteria Rating Systems", IEEE Intelligent Systems, Vol.22, No.3(2007), 48-55.
Aggarwal, C. C., C. Procopiuc and P. S. Yu, "Finding Localized Associations in Market Basket Data", IEEE Transactions on Knowledge and Data Engineering, Vol.14, No.1 (2002), 51-62.
Ahn, H. J., "A New Similarity Measure for Collaborative Filtering to Alleviate the New User ColdStarting Problem", Information Sciences, Vol.178, No.1(2008), 37-51.
Albadvi, A. and M. Shahbazi, "A Hybrid Recommendation Technique based on Product Category Attributes", Expert Systems with Applications, Vol.36, No.9(2009), 11480-11488.
BaezaYates, R. A. and B. RibeiroNeto, Modern Information Retrieval, AddisonWesley Longman Publishing Co., Inc., Boston, 1999.
Balabanovic, M. and Y. Shoham, "ContentBased, Collaborative Recommendation", Communications of the ACM, Vol.40, No.3(1998), 66-72.
Belkin, N. J. and W. B. Croft, "Information Filtering and Information Retrieval: Two Sides of the Same Coin", Communications of the ACM, Vol.35, No.12(1992), 29-38.
Billsus, D. and M. J. Pazzani, "Learning Collaborative Information Filters", 15th International Conference on Machine Learning, (1998), 48-56
Billsus, D. and M. J. Pazzani, "User Modeling for Adaptive News Access", User Modeling and UserAdapted Interaction, Vol.10, No.2 (2000), 147-180.
Bobadilla, J., F. Serradilla and J. Bernal, "A New Collaborative Filtering Metric that Improves the Behavior of Recommender Systems", KnowledgeBased Systems, Vol.23, No.6(2010), 520-528.
Chen, L. S., F. H. Hsu, M. C. Chen and Y. C. Hsu, "Developing Recommender Systems with the Consideration of Product Profitability for Sellers", Information Sciences, Vol.178, No.4(2008), 1032-1048.
Cheung, K. W., J. T. Kwok, M. H. Law, and K. C. Tsui, "Mining Customer Product Ratings for Personalized Marketing", Decision Support Systems, Vol.35, No.2(2003), 231-243.
Getoor, L. and M. Sahami, "Using Probabilistic Relational Models for Collaborative Filtering", Workshop on Web Usage Analysis and User Profiling, (1999), 1-6.
Goldberg, D., D. Nichols, B. M. Oki and D. Terry, "Using Collabortive Filtering to Weave an Information Tapestry", Communications of the ACM, Vol.35, No.12(1992), 61-70.
Goldberg, K., T. Roeder, D. Gupta and C. Perkins, "Eigentaste: A Constant Time Collaborative Filtering Algorithm", Information Retrieval, Vol.4, No.2(2001), 133-151.
Hofmann, T., "Collaborative Filtering via Gaussian Probabilistic Latent Semantic Analysis", 26th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, (2002), 259-266.
Hofmann, T., "Latent Semantic Models for Collaborative Filtering", ACM Transactions on Information Systems, Vol.22, No.1(2004), 89-115.
Huang, C. L. and W. L. Huang, "Handling Sequential Pattern Decay: Developing a Two Stage Collaborative Recommendation System", Electronic Commerce Research and Applications, Vol.8, No.3(2009), 117-129.
Jeong, B., J. Lee and H. Cho, "An Iterative SemiExplicit Rating Method for Building Collaborative Recommender Systems", Expert Systems with Applications, Vol.36, No.3 (2009), 6181-6186.
Joaquin, D. and I. Naohiro, "Memorybased WeightedMajority Prediction", ACM SIGIR ？99 Workshop on Recommender Systems: Algorithms and Evaluation, (1999), 1-5.
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.
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.
Kim, J. K., H. K. Kim and Y. H. Cho, "A UserOriented Contents Recommendation System in PeertoPeer Architecture", Expert Systems with Applications, Vol.34, No.1(2008), 300-312.
Kumar, R., P. Raghavan, S. Rajagopalan and A. Tomkins, "Recommender Systems: A Probabilistic Analysis", Journal of Computer and System Science, Vol.63, No.1(2001), 42-61.
Kwon, K., J. Cho and Y. Park, "Multidimensional Credibility Model for Neighbor Selection in Collaborative Recommendation", Expert Systems with Applications, Vol.36, No.3 (2009), 7114-7122.
Lang, K., "NewsWeeder: Learning to Filter Netnews", 12th International Conference on Machine Learning, (1995), 1-9.
Lee, J. S. and S. Olafsson, "Twoway Cooperative Prediction for Collaborative Filtering Recommendations", Expert Systems with Applications, Vol.36, No.3(2009), 5353-5361.
Lee, T. Q., Y. Park and Y. T. Park, "A Timebased Approach to Effective Recommender Systems Using Implicit Feedback", Expert Systems with Applications, Vol.34, No.4 (2008), 3055-3062.
Liu, D. R., C. H. Lai and W. J. Lee, "A Hybrid of Sequential Rules and Collaborative Filtering for Product Recommendation", Information Sciences, Vol.179, No.20(2009), 3505-3519.
Liu Z., W. Qu, H. Li and C. Xie, "A Hybrid Collaborative Filtering Recommendation Mechanism for P2P Networks", Future Generation Computer Systems, Vol.26, No.8(2010), 1409-1417.
Mooney, R. J. and L. Roy, "ContentBased Book Recommending Using Learning for Text Categorization", ACM SIGIR ？ 99 Workshop on Recommender Systems: Algorithms and Evaluation, (1999), 1-8.
Nakamura, A. and N. Abe, "Collaborative Filtering using Weighted Majority Prediction Algorithm", 15th International Conference on Machine Learning, (1998), 395-403.
Park, Y. J. and K. N. Chang, "Individual and Group Behaviorbased Customer Profile Model for Personalized Product Recommendation", Expert Systems with Applications, Vol.36, No.2(2009), 1932-1939.
Pavlov, D. Y. and D. M. Pennock, "A Maximum Entropy Approach to Collaborative Filtering in Dynamic, Sparse, HighDimensional Domains", Advances in Neural Information Processing Systems, Vol.15(2002), 1441-1448.
Pazzani, M. and D. Billsus, "Learning and Revising User Profile: The Identification of Interesting Web Sites", Machine Learning, Vol.27, No.3(1997), 313-331.
Pennock, D. M., E. Horvitz, S. Lawrence and C. L. Giles, "Collaborative Filtering by Personality Diagnosis: A Hybrid Memoryand Model Based Approach", 16th Conference on Uncertainty in Artificial Intelligence, (1999), 473-480.
Resnick, P., N. Iacovou, M. Suchak, P. Bergstrom and J. Riedl, "GroupLens: An Open Architecture for Collaborative Filtering of Netnews", ACM Conference on Computer Supported Cooperative Work, (1994), 175-186.
Russell, S. and V. Yoon, "Applications of Wavelet Data Reduction in a Recommender System", Expert Systems with Applications, Vol. 34, No.4(2008), 2316-2325.
Salter, J. and N. Antonopoulos, "Cinema Screen Recommender Agent: Combining Collaborative and ContentBased Filtering", IEEE Intelligent Systems, Vol.21, No.1(2006), 35-41.
Shani, G., D. Heckerman, and R. I. Brafman, "An MDPbased Recommender System", Journal of Machine Learning Research, Vol.6, No.2 (2002), 1265-1295.
Shardanand, U. and P. Maes, "Social Information Filtering Algorithms for Automating 'Word of Mouth'", SIGCHI Conference on Human Factors in Computing Systems, (1995), 210-217.
Si, L. and R. Jin, "Flexible Mixture Model for Collaborative Filtering", 20th International Conference on Machine Learning, (2003), 1-8.
Symeonidis, P., A. Nanopoulos and Y. Manolopoulos, "Providing Justifications in Recommender Systems", IEEE Transactions on Systems, Man and Cybernetics-Part A: Systems an Humans, Vol.38, No.6(2008), 1262-1272.
Tang, T. Y., P. Winoto and K. C. C. Chan, "Scaling Down Candidate Sets based on the Temporal Feature of Items for Improved Hybrid Recommendations", Intelligent Techniques for Web Personalization, (2005), 169-186.
Wang, Y., W. Dai and Y. Yuan, "Website Browsing Aid: A Navigation Graphbased Recommendation System", Decision Support Systems, Vol.45, No.3(2008), 387-400.
Wei, C. P., C. S. Yang and H. W. Hsiao, "A Collaborative Filteringbased Approach to Personalized Document Clustering", Decision Support Systems, Vol.45, No.3(2008), 413-428.
Yu, K., A. Schwaighofer, V. Tresp, X. Xu and H. P. Kriegel, "Probabilistic Memorybased Collaborative Filtering", IEEE Transactions on Knowledge and Data Engineering, Vol. 16, No.1(2004), 56-69.
Yu, K., X. Xu, J. Tao, M. Ester, and H. P. Kriegel, "Instance Selection Techniques for Memorybased Collaborative Filtering", 2nd SIAM International Conference on Data Mining, (2002), 1-16.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.