Information Filtering via Improved Similarity Definition
PAN Xin1, DENG Gui-Shi1, LIU Jian-Guo2,3
1Institute of Systems Engineering, Dalian University of Technology, Dalian 116023 2Research Center of Complex Systems Science, Shanghai University of Science and Technology, Shanghai 200093 3Business School, Shanghai University of Science and Technology, Shanghai 200093
Information Filtering via Improved Similarity Definition
PAN Xin1, DENG Gui-Shi1, LIU Jian-Guo2,3
1Institute of Systems Engineering, Dalian University of Technology, Dalian 116023 2Research Center of Complex Systems Science, Shanghai University of Science and Technology, Shanghai 200093 3Business School, Shanghai University of Science and Technology, Shanghai 200093
Based on a new definition of user similarity, we introduce an improved collaborative filtering (ICF) algorithm, which could improve the algorithmic accuracy and diversity simultaneously. In the ICF, instead of the standard Pearson coefficient, the user-user similarities are obtained by integrating the heat conduction and mass diffusion processes. The simulation results on a benchmark data set indicate that the corresponding algorithmic accuracy, measured by the ranking score, is improved by 6.7% in the optimal case compared to the standard collaborative filtering (CF) algorithm. More importantly, the diversity of the recommendation lists is also improved by 63.6%. Since the user similarity is crucial for the CF algorithm, this work may shed some light on how to improve the algorithmic performance by giving accurate similarity measurement.
Based on a new definition of user similarity, we introduce an improved collaborative filtering (ICF) algorithm, which could improve the algorithmic accuracy and diversity simultaneously. In the ICF, instead of the standard Pearson coefficient, the user-user similarities are obtained by integrating the heat conduction and mass diffusion processes. The simulation results on a benchmark data set indicate that the corresponding algorithmic accuracy, measured by the ranking score, is improved by 6.7% in the optimal case compared to the standard collaborative filtering (CF) algorithm. More importantly, the diversity of the recommendation lists is also improved by 63.6%. Since the user similarity is crucial for the CF algorithm, this work may shed some light on how to improve the algorithmic performance by giving accurate similarity measurement.
PAN Xin;DENG Gui-Shi;LIU Jian-Guo;. Information Filtering via Improved Similarity Definition[J]. 中国物理快报, 2010, 27(6): 68903-068903.
PAN Xin, DENG Gui-Shi, LIU Jian-Guo,. Information Filtering via Improved Similarity Definition. Chin. Phys. Lett., 2010, 27(6): 68903-068903.
[1] Zhang G Q, Zhang G Q, Yang Q F, Cheng S Q, and Zhou T 2008 New J. Phys. 10 12307 [2] Resnick P, Varian H R 1997 Commun. ACM 40 56 [3] Adomavicius G and Tuzhilin A 2005 IEEE Trans. Know. Data Eng. 17 734 [4] Herlocker J L, Konstan J A, Terveen K and Riedl J 2004 ACM Trans. Inform. Syst. 22 5 [5] Konstan J A, Miller B N, Maltz D, Herlocker J L, Gordon L R and Riedl J 1997 Commun. ACM 40 77 [6] Liu J G, Wang B H, Guo Q 2009 Int. J. Mod. Phys. C 20 285 [7] Liu J G, Zhou T, Che H A, Wang B H and Zhang Y C 2010 Physica A 389 881 [8] Liu J G, Zhou T, Wang B H, Zhang Y C and Guo Q 2010 J. Mod. Phys. C 20 1925 [9] Liu R R, Jia C X, Zhou T, Sun D and Wang B H 2009 Physica A 388 462 [10] Sun D, Zhou T, Liu J G, Liu R R, Jia C X and Wang B H 2009 Phys. Rev. E 80 017101 [11] Balabanovic M and Shoham Y 1997 Commun. ACM 40 66 [12] Pazzani M J 1999 Artif. Intell. Rev. 13 393 [13] Gao Y, Luo H and Fan J 2009 Lect. Notes Comput. Sci. 5371 217 [14] Luo H, Fan J, Keim D A and Satoh S 2009 Lect. Notes Comput. Sci. 5371 459 [15] Pazzani M and Billsus D 1997 Machine Learning 27 313 [16] Basu C, Hirsh H and Cohen W 1998 Technical Report WS-98-08 (New York: AAAI Press) p 714 [17] Good N, Schafer J B, Konstan J A, Borchers A L, Sarwar B, Herlocker J and Riedl J 1999 Proc. Conf. Am. Assoc. Artif. Intell. p 439 [18] Zhang Y C, Medo M, Ren J, Zhou T, Li T and Yang F 2008 Europhys. Lett. 80 68003 [19] Zhou T, Ren J, Medo M and Zhang Y C 2007 Phys. Rev. E 76 046115 [20] Zhang Y C, Blattner M and Yu Y K 2007 Phys. Rev. Lett. 99 154301 [21] Zhou T, Jiang L L, Su R Q and Zhang Y C 2008 Europhys. Lett. 81 58004 [22] Zhou T, Su R Q, Liu R R, Jiang L L, Wang B H and Zhang Y C 2009 New J Physics 11 123008 [23] Zhou T, Kuscsik Z, Liu J G, Medo M, Wakeling J R, and Zhang Y C 2010 Proc. Natl. Acad. Sci. U.S.A. 107 4511 [24] Ou Q, Jin Y D, Zhou T, Wang B H and Yin B Q 2007 Phys. Rev. E 75 021102 [25] Lind P G, González M C and Herrmann H J 2006 Phys. Rev. E 72 056127 [26] Lind P G and Herrmann H J 2007 New J. Phys. 9 228 [27] Liu J G, Dang Y Z and Wang Z T 2005 Mod. Phys. Lett. B 19 785 [28] Liu J G, Dang Y Z and Wang Z T 2006 Mod. Phys. Lett. B 20 815 [29] Liu J G, Dang Y Z and Wang Z T 2006 Physica A 366 578 [30] Liu J G, Xuan Z G, Dang Y Z, Guo Q and Wang Z T 2007 Physica A 377 302 [31] Zhou T, Lü L and Zhang Y C 2009 Eur. Phys. J. B 71 623
2010, Vol. 27 
