Method and Apparatus for Deep Learning Based Real-time Program Recommendation

摘要

A real-time program recommendation method based on deep learning is provided. The method comprises the steps of: receiving customer information of a specific customer and information about a plurality of programs - The customer information of the specific customer includes real-time program viewing sequence information, and the plurality of programs are selected by the specific customer. Programs currently being aired on channels other than the channel - Providing a viewing probability value for each of the plurality of programs based on customer information of the specific customer and information on the plurality of programs - The viewing probability value is indicating a probability that the specific customer will watch the corresponding program; and determining to recommend at least one program among the plurality of programs to the specific customer based on the viewing probability values.