A Two-stage Recommendation Algorithm Based on K-means Clustering In Mobile E-commerce
Fuzhi Zhang, Huilin Liu, Jinbo Chao · 2010
In the mobile e-commerce, location-based recommendation algorithm is widely adopted in location-based personalized information services, which considers the impact of location for the recommendation results. But its accuracy is poor if it only uses the location as the primary standard of the recommendation result or selects the neighbor users through the positional similarity of the mobile users. To improve the quality of recommendation, we present a two-stage recommendation algorithm based on K-means clustering (TSRA). Neighbor users are selected according to the location of users. K-means clustering algorithm is used to cluster the profile of the neighbor users. According to the difference of location-sensitive of the active users, we use collaborative filtering algorithm by integrating location sensitivity. Experimental results show that the proposed algorithm has a higher accuracy.