Personalized E-Commerce Recommendation Based on Ontology
Peiguang Lin, Feng Yang, Xiao Jin Yu, Qun Jie Xu · 2008
The current collaborative recommendation approaches mainly measure users' similarity by comparing user's entire interests and don't consider user's interest quality, especially interest span. With so many goods in the E-commerce web site, how to get the needed product quickly so as to promote the efficiency of E-commerce system? This paper presented a personalized recommendation method based on ontology. To improve the precision, we firstly divided users' interests into long-time interests and short-time interests; and then by use of the principle of partial similarity, the recommendation mechanism and algorithm were given. Lastly, based on the method above, a prototype system was presented and the system test was done. Experimental results indicate that this method can recommend related products in the majority to target users and it can be practical.