Personal TV content recommendation list generating method

摘要

Present invention explains a channel recommendation list method for IPTV applications. In said method, two main algorithms are used to generate recommendation list. One of said algorithms is based on users watching preferences according to present and past experiences. In described algorithm, users are clustered into groups according to present and past experiences. Moreover, the other algorithm is based on user's watching behaviour. In described algorithm, watching frequency of a channel for sampled days, for each user is calculated. By combining the results of said algorithms, a successful recommendation list is generated.