Improved collaborative filtering algorithm based
on heat conduction

doi:10.1007/s11704-009-0050-2

PDF(301 KB)

Front. Comput. Sci. ›› 2009, Vol. 3 ›› Issue (3) : 417-420. DOI: 10.1007/s11704-009-0050-2

Research articles

Improved collaborative filtering algorithm based on heat conduction

Qiang GUO ¹, Jianguo LIU ², Binghong WANG ³,

Author information +

History +

Abstract

In this paper, we present an improved collaborative filtering (ICF) algorithm by using the heat diffusion process to generate the user correlation. This algorithm has remarkably higher accuracy than the standard collaborative filtering (CF) using Pearson correlation. Furthermore, we introduce a free parameter β to regulate the contributions of objects to user correlation. The numerical simulation results indicate that decreasing the influence of popular objects can further improve the algorithmic accuracy and diversity.

Keywords

recommendation algorithm / collaborative filtering / heat conduction

Cite this article

EndNote

Ris (Procite)

Bibtex

Download citation ▾

Qiang GUO , Jianguo LIU , Binghong WANG ,. Improved collaborative filtering algorithm based on heat conduction. Front. Comput. Sci., 2009, 3(3): 417‒420 https://doi.org/10.1007/s11704-009-0050-2