MAP
Figure 4.6: MAP of proposed systems with different iterations.
User−based CF Item−based CF Content−based Initial 3 Iterations 4 Iterations 8 Iterations 0
0.01 0.02 0.03 0.04 0.05 0.06 0.07
MAP
Figure 4.7: The comparison of MAP.
Chapter 5 Conclusion
This paper proposes a recommendation system (RS) with two concept spaces, item-based and user-item-based concepts, in an iterative procedure. The actual groups of ar-ticles are named item-based concepts and the expected groups of arar-ticles are named user-based concepts. The user-based concepts are reinforced from item-based con-cepts with readers’ reading behaviors. When two articles are both read by one or more users, the two articles are related. The more users read both of them, the stronger the relations. A series of keywords are generated from the contents of ar-ticles and used as the dimension to form item-based concepts in the next iteration.
Reading dependence is also considered when generating recommendation. We place both readers and articles in concept space and compute the association. In our case study, the average precision-recall curves indicate that the proposed RS produces more hits with more iterations. In the perspective of average MAP, we find that it generally increases. To some extent, it is higher than user-based/item-based CF and content-based filtering.
Our proposed RS deal with scalability and sparsity with clustering methods.
Each cluster is a group of articles and named concept. Our proposed RS can dy-namically scale the scale of clusters up and down in the iterative procedure. The size increases as necessary and merely increase to 87 from 5 while the dimension of user-based CF is 3, 200 and 30, 000 for item-user-based CF. In other hand, the proposed RS cluster articles with keywords as dimension. Thus the new articles can be clustered and recommended. The size of keywords increases as necessary in each iteration to separated concepts. According to the results of PR curves and average MAP, the iterative procedure can further increase the accuracy of the recommendation.
The proposed RS can be applied to other field in such a manner. Just as we extract keywords from textual articles, we can extract video and audio features to apply the RS in multimedia recommendations. As long as the items to be
recom-mended contain extractable features, we believe that the idea of reinforcing item-based concepts to user-item-based concepts can be extended to any situation with regard to interactions between users and items.
Bibliography
[1] C. Basu, H. Hirsh, and W. Cohen. Recommendation as classification: Using social and content-based information in recommendation. In In Proceedings of the Fifteenth National Conference on Artificial Intelligence, pages 714–720.
AAAI Press, 1998.
[2] J. S. Breese, D. Heckerman, and C. Kadie. Empirical analysis of predictive al-gorithms for collaborative filtering. In Proceedings of the Fourteenth Conference on Uncertainty in Artificial Intelligence, UAI’98, pages 43–52, San Francisco, CA, USA, 1998. Morgan Kaufmann Publishers Inc.
[3] M. Claypool, A. Gokhale, T. Miranda, P. Murnikov, D. Netes, and M. Sartin.
Combining content-based and collaborative filters in an online newspaper. In Proceedings of ACM SIGIR workshop on recommender systems, volume 60.
Citeseer, 1999.
[4] M. Connor and J. Herlocker. Clustering items for collaborative filtering, 2001.
[5] J. Davidson, B. Liebald, J. Liu, P. Nandy, T. Van Vleet, U. Gargi, S. Gupta, Y. He, M. Lambert, B. Livingston, and D. Sampath. The youtube video rec-ommendation system. In Proceedings of the Fourth ACM Conference on Rec-ommender Systems, RecSys ’10, pages 293–296, New York, NY, USA, 2010.
ACM.
[6] J. Davis and M. Goadrich. The relationship between precision-recall and roc curves. In Proceedings of the 23rd international conference on Machine learning, pages 233–240. ACM, 2006.
[7] S. C. Deerwester, S. T. Dumais, T. K. Landauer, G. W. Furnas, and R. A.
Harshman. Indexing by latent semantic analysis. JASIS, 41(6):391–407, 1990.
[8] J. A. Hartigan. Clustering. Annual review of biophysics and bioengineering, 2(1):81–102, 1973.
[9] J. A. Hartigan and M. A. Wong. Algorithm as 136: A k-means clustering algorithm. Applied statistics, pages 100–108, 1979.
[10] J. L. Herlocker, J. A. Konstan, A. Borchers, and J. Riedl. An algorithmic framework for performing collaborative filtering. In Proceedings of the 22Nd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR ’99, pages 230–237, New York, NY, USA, 1999.
ACM.
[11] J. L. Herlocker, J. A. Konstan, A. Borchers, and J. Riedl. An algorithmic framework for performing collaborative filtering. In Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information retrieval, pages 230–237. ACM, 1999.
[12] F. Jelinek. Statistical methods for speech recognition. 1997.
[13] J. A. Konstan, B. N. Miller, D. Maltz, J. L. Herlocker, L. R. Gordon, and J. Riedl. Grouplens: applying collaborative filtering to usenet news. Commu-nications of the ACM, 40(3):77–87, 1997.
[14] G. Linden, J. Jacobi, and E. Benson. Collaborative recommendations using item-to-item similarity mappings, July 24 2001. US Patent 6,266,649.
[15] G. Linden, B. Smith, and J. York. Amazon.com recommendations: item-to-item collaborative filtering. Internet Computing, IEEE, 7(1):76–80, Jan 2003.
[16] H. T. Ng, W. B. Goh, and K. L. Low. Feature selection, perceptron learning, and a usability case study for text categorization. SIGIR Forum, 31(SI):67–73, July 1997.
[17] G. Salton, A. Wong, and C. S. Yang. A vector space model for automatic indexing. Commun. ACM, 18(11):613–620, Nov. 1975.
[18] B. Sarwar, G. Karypis, J. Konstan, and J. Riedl. Analysis of recommendation algorithms for e-commerce. In Proceedings of the 2Nd ACM Conference on Electronic Commerce, EC ’00, pages 158–167, New York, NY, USA, 2000. ACM.
[19] B. Sarwar, G. Karypis, J. Konstan, and J. Riedl. Item-based collaborative filtering recommendation algorithms. In Proceedings of the 10th International Conference on World Wide Web, WWW ’01, pages 285–295, New York, NY, USA, 2001. ACM.
[20] F. Sebastiani. Machine learning in automated text categorization. ACM Com-put. Surv., 34(1):1–47, Mar. 2002.
[21] G. Shani and A. Gunawardana. Evaluating recommendation systems. In Rec-ommender systems handbook, pages 257–297. Springer, 2011.
[22] A. Singhal. Modern information retrieval: A brief overview. Bulletin of the IEEE Computer Society Technical Committee on Data Engineering, 24(4):35–
43, 2001.
[23] X. Su and T. M. Khoshgoftaar. A survey of collaborative filtering techniques.
Advances in artificial intelligence, 2009:4, 2009.
[24] X. Su and T. M. Khoshgoftaar. A survey of collaborative filtering techniques.
Adv. in Artif. Intell., 2009:4:2–4:2, Jan. 2009.
[25] L. Terveen, W. Hill, B. Amento, D. McDonald, and J. Creter. Phoaks: A system for sharing recommendations. Communications of the ACM, 40(3):59–62, 1997.
[26] H. Tseng. A conditional random field word segmenter. In In Fourth SIGHAN Workshop on Chinese Language Processing, 2005.
[27] L. Ungar, D. Foster, E. Andre, S. Wars, F. S. Wars, D. S. Wars, and J. H.
Whispers. Clustering methods for collaborative filtering. AAAI Press, 1998.
[28] R. Van Meteren and M. Van Someren. Using content-based filtering for rec-ommendation. In Proceedings of the Machine Learning in the New Information Age: MLnet/ECML2000 Workshop, 2000.
[29] C. Wang and D. M. Blei. Collaborative topic modeling for recommending scientific articles. In Proceedings of the 17th ACM SIGKDD International Con-ference on Knowledge Discovery and Data Mining, KDD ’11, pages 448–456, New York, NY, USA, 2011. ACM.