EFFECTS OF BINARY SIMILARITY MEASURES ON TOP-N RECOMMENDATIONS

Year 2013, Volume: 14 Issue: 1, 55 - 65, 03.10.2013

Edip Şenyürek Hüseyin Polat

Abstract

Shopping over the Internet through several e-commerce sites is receiving increasing attention. Customers want to purchase those products that they might like without wasting time and/or money. To help their customers, many online companies provide top-N recommendations by means of recommender systems. Similarity measures used to find out the most similar entities might affect the
overall performance of top-N predictions. Although there are various binary ratings-based similarity metrics, their effects on accuracy and online efficiency of top-N recommendations have not been deeply studied.
In this study, we investigate seven well-known binary ratings-based similarity metrics in terms of both preciseness and efficiency while providing top-N recommendations. To compare them with respect to accuracy and competence, we perform several experiments based on two well-known real data sets. We modify top-N recommendation algorithm in such a way so that the most similar users' data
are involved in recommendation process. We also study how varying controlling parameters affect overall performance with different similarity metrics. We analyze our empirical results and provide some suggestions.

Keywords

Similarity metric, Top-N recommendation, Accuracy, Online efficiency

Uygulamalı Bilimler ve Mühendislik

Abstract

İnternet üzerinden sanal firmalar aracılığıyla alışveriş yapmak artan ilgi görmektedir. Müşteriler beğenebilecekleri ürünleri zaman ve/veya paralarını boşa harcamadan satın almak isterler. Müşterilerine bu süreçte yardımcı olmak için birçok sanal şirket öneri sistemlerinden yararlanıp müşterilerine en-iyi-N önerileri sunmaktadır. En benzer varlıkları belirlemede kullanılan benzerlik ölçütleri en-iyi-N önerileri hizmetinin genel performansını etkileyebilir. İkili değerler üzerinde işlem yapan birçok benzerlik ölçütü bulunmasına rağmen bunların en-iyi-N önerilerinin doğruluğu ve çevrimiçi performansı üzerindeki etkisi detaylı biçimde çalışılmamıştır. yapıldı. Ayrıca en-iyi-N öneri algoritması en benzer kullanıcıların verisi öneri üretilirken kullanılacak şekilde değiştirildi. Değişen kontrol parametrelerinin performansa olan etkisi araştırıldı. Deneysel sonuçlar doğruluk ve performans açısından analiz edilerek bazı öneriler sunuldu

Keywords

Benzerlik ölçütü, En-iyi-N önerisi, Doğruluk, Çevrimiçi performans

References

• Billus, D. and Pazzani, M.J. (1998). Learning Collaborative Information Filters. Proceed- ings of the 15th International Conference on Machine Learning, Adison, WI, USA, 46- 54.
• Brožovský, L. (2006). Recommender System for A Dating Service. Master’s thesis, Prague, Czech Republic: Charles University in Prague.
• Cha, S.-H., Yoon, S., and Tappert, C.C. (2005). On Binary Similarity Measures for Hand- written Character Recognition. Proceedings of the 2005 8th International Conference on Document Analysis and Recognition, Seoul, Korea, vol. 1, 4-8.
• Choi, S.-S., Cha, S.-H., and Tappert, C.C. (2010). A Survey of Binary Similarity and Distance Measures. Journal of Systemics, Cybernetics and Informatics 8 (1), 43-48.
• GroupLens Research, Data Sets, 2006. http://www.grouplens.org/node/12. (May 15, 2012).
• Kaleli, C. and Polat, H. (2009). Similar or Dis- similar Users? Or both? Proceedings of the 2009 2nd International Symposium on Elec- tronic Commerce and Security, Nanchang City, China, vol. 2, 184-189.
• Miranda, C. and Jorge, A.M. (2009). Item-based and user-based incremental collaborative fil- tering for Web recommendations. Lecture Notes in Computer Science 5816, 673-684.
• Miyahara, K. and Pazzani, M.J. (2000). Collabo- rative filtering with the simple Bayesian clas- sifier. Lecture Notes in Computer Science 1886, 679-689.
• Miyahara, K. and Pazzani, M.J. (2002). Im- provement collaborative filtering with the simple Bayesian classifier. Transactions of Information Processing Society of Japan 43 (11), 3429-3437.
• Papagelis, M., Rousidis, I., Plexousakis, D., and Theoharopoulos, E. (2005). Incremental Col- laborative Filtering for Highly-Scalable Rec- ommendation Algorithms. Lecture Notes in Computer Science 3488, 553-561.
• Resnick, P., Iacovou, N., Suchak, M., Bergstrom, P., and Riedl, J.T. (1994). GroupLens: An Open Architecture for Collaborative Filtering of Netnews. Proceedings of the 1994 ACM Conference on Computer Supported Cooper- ative Work, Chapel Hill, NC, USA, 175-186.
• Robu, V. and Poutré, H.L. (2009). Learning The Structure of Utility Graphs Used In Multi- Issue Negotiation Through Collaborative Fil- tering. Lecture Notes in Computer Science 4078, 192-209.
• Sarwar, B.M., Karypis, G., Konstan, J.A., and Riedl, J.T. (2001). Item-Based Collaborative Filtering Recommendation Algorithms. Pro- ceedings of the 10th International Confer- ence on World Wide Web, Hong Kong, 285- 295.
• Stata Corp. LP, Stata 11 help for measure op- tion, 1996. http://www.stata.com/help.cgi?measure+opti on (May 15 2012).
• Tubbs, J.D. (1989). A Note on Binary Template Matching. Pattern Recognition 22 (4), 359- 365. Veal, B.W.G. (2011). Binary Similarity Measures And Their Applications In Machine Learning. PhD thesis, London, United King- dom: London School of Economics.
• Vozalis, M. and Margaritis, K.G. (2004). Unison-CF: A Multiple-Component, Adap- tive Collaborative Filtering System. Lecture Notes in Computer Science 3137, 255-264.
• Yang, C.C., Chen, H., and Hong, K. (2003). Visualization of Large Category Map For In- ternet Browsing. Decision Support Systems 35 (1), 89-102.
• Zhang, B. and Srihari, S.N. (2003). Binary Vec- tor Dissimilarity Measures for Handwriting Identification. Document Recognition and Retrieval X vol. 5010 (1), 28-38.