A research of recommendation algorithm based on cloud model

With the popularity of Internet and the development of e-commerce, recommendation system is becoming more and more important research content of e-commerce IT technology. The problem is the fuzziness and uncertainty of traditional recommendation system that caused by extremely sparse rating data. The paper use cloud model which are knowledge representation in quality and bridge function of the conversion between quality and quantity, provide an item classification recommendation algorithm based on cloud model. The algorithm takes fully into account the classification of specific items and users´ different favorable levels to the same rating or not rating item, avoids the shortcoming of treating user totally rating as single vector, and can overcome effectively the affects of extremely sparse user ratings. The experiment result show the performance of item classification recommendation algorithm based on cloud model is better.