Collaborative filtering recommendation algorithm based on look-ahead selective sampling

Personalized Recommendation System has become an important research item to prove the suitable product and services for individual. And classification of customers becomes the basis to produce recommendation. In a realistic EC system, the magnitudes of customers and products are all huge, so the quality of recommendation decreases dramatically. To improve recommending quantity, a collaborative filtering model was proposed based on look-ahead sampling. In n-dimension Euclid space constituted by users, the proposed algorithm reduces the number of samples while maintaining the quality of classification, through estimating sample’s utility for classifier. At last, experiments were designed at the basis of MoveLens dataset. Compared with general collaborative filtering, the proposed algorithm has higher quality of recommendation.