Diffusion-Based Recommendation in Collaborative Tagging Systems
SHANG Ming-Sheng1, ZHANG Zi-Ke2
1School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 6100542Department of Physics, University of Fribourg, CH-1700 Fribourg, Switzerland
Diffusion-Based Recommendation in Collaborative Tagging Systems
SHANG Ming-Sheng1, ZHANG Zi-Ke2
1School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 6100542Department of Physics, University of Fribourg, CH-1700 Fribourg, Switzerland
摘要Recently, collaborative tagging systems have attracted more and more attention and have been widely applied in web systems. Tags provide highly abstracted information about personal preferences and item content, and therefore have the potential to help in improving better personalized recommendations. We propose a diffusion-based recommendation algorithm considering the personal vocabulary and evaluate it in a real-world dataset: Del.icio.us. Experimental results demonstrate that the usage of tag information can significantly improve the accuracy of personalized recommendations.
Abstract：Recently, collaborative tagging systems have attracted more and more attention and have been widely applied in web systems. Tags provide highly abstracted information about personal preferences and item content, and therefore have the potential to help in improving better personalized recommendations. We propose a diffusion-based recommendation algorithm considering the personal vocabulary and evaluate it in a real-world dataset: Del.icio.us. Experimental results demonstrate that the usage of tag information can significantly improve the accuracy of personalized recommendations.
[1] Brin S and Page L 1998 Comput. Netw. ISDN Syst. 30 107 [2] Resnick P and Varian H R 1997 Commun. ACM 4056 [3] Kazienko P et al 2007 Inf. Sci. 177 2269 [4] Pazzani M J and Billsus D 2007 LNCS 4321 325 [5] Tso K and Schmidt-Thieme L 2005 Proc. 29th AnnualConference of the German Classification Society (Magdeburg, Germany 9--11 March 2005) [6] Hotho A et al. 2006 LNCS 4011 411 [7] Cattuto C et al 2008 PNAS 104 1461 [8] Golder S A and Huberman B A 2006 J. Inf. Sci. 32 198 [9] Mishne G 2006 Proc. $15^{\rm th$ WWW (Edinburgh, Scotland 23--26 May 2006) ) p 953 [10] Sigurbj\"{ornsson B and Zwol R V 2008 Proc. $17^{\rm th$ WWW (Beijing, China 21--25 April 2008) p 327 [11] Schenkel R et al 2008 Proc. $31^{\rm th$ ACMSIGIR (Singapore 20--24 July 2008) p 523 [12] Nakamoto R Y et al 2007 IAENG Int. J. Comput. Sci. 34 214 [13] Tso-Sutter K H L et al 2008 Proc. $23^{\rm rd$ACM SAC 2008 (Fortaleza, Cear\'a, Brazil 16--20 March 2008) p 1995 [14] Chi Y et al 2008 Proc. 17th ACM Conference onInformation and Knowledge Management (California 26--30 October 2008) p 941 [15] Shepitsen A et al 2008 Proc. ACM Conference onRecommender Systems (Lausanne, Switzerland 23--25 October 2008) p 259 [16] Zhang Z K et al 2009 Physica A (accepted) [17] Shang M S et al 2009 Physica A 388 4867 [18] Zhou T et al 2006 Prog. Nat. Sci. 16 452 [19] Zhou Y Z et al 2007 Chin. Phys. Lett. 24 581 [20] Zhao H and Gao Z Y 2007 Chin. Phys. Lett. 241114 [21] Zhang H F et al 2009 Chin. Phys. Lett. 26068901 [22] Yan G et al 2005 Chin. Phys. Lett. 22 510 [23] Zhang Y C et al 2007 Phys. Rev. Lett. 99154301 [24] Zhang Y C et al 2007 Eur. Phys. Lett. 8068003 [25] Zhou T et al 2007 Phys. Rev. E 76 046115 [26] Zhou T et al 2008 Eur. Phys. Lett. 81 58004 [27] Herlocker J L et al 2004 ACM Trans. Inf. Syst. 22 5 [28] Zhang Z K et al 2008 Eur. Phys. J. B 66 557 [29] Ghoshal G et al 2009 Phys. Rev. E 79 066118 [30] Shi J J et al 2009 Chin. Phys. Lett. 26078902 [31] Fu C H et al 2008 Chin. Phys. Lett. 25 4181
2009, Vol. 26 
