A Data Filling Algorithm Based on Cloud Model
QI Yu-feng · Computer Technology and Development · 2010
The collaborative filtering is one of the most important technologies in current E-commerce system,in the view of data in collaborative filtering technology are extremely sparse resulting in bad similarity measure and recommend poor quality,using a cloud model'action in qualitative knowledge representation and the role of conversion among qualitative and quantitatve knowledge,proposed a data filling algorithm based on cloud model,and using the classical experimental data to validate and compare,it calculate user score missing items by using similar user.The result shows,even if the user's rating data is extremely sparse,it can get better recommendation quality by filling data with the algorithm and adopting traditional collaborative filtering algorithm,to some extent it can solve common sparse problems in recommended system.