TV program recommendation method

Abstract

The invention requests to protect a recommendation method of a TV program. The method comprises the following steps: firstly receiving language information input by a user; obtaining a characteristicelement according to the language information, retrieving and evaluating the electronic program information in an electronic program table database by taking the characteristic element as the retrieval keyword, and extracting the related electronic program information according to the similarity; constructing a statistical model by utilizing the characteristic set and the machine learning method;matching the programs in the electronic program table database by using the statistical model; outputting a matching result to the user; secondly, sending the program request information by the user,acquiring score information on the program by the user to obtain a user-program score matrix, and initially recommending the program; if the initially recommended program is half same as the matchingresult, recommending based on this situation, if the initially recommended program is different from the matching result, selecting the program with the second-highest score as the target program, andrepeating the above steps to perform the filling, and recommending the program list to the user until more than half recommended programs are same as the matching result. The accuracy of the programrecommendation can be improved through the mixed recommendation method.

Description

technical field [0001] The invention belongs to the technical field of recommendation, in particular to a method for recommending television programs. Background technique [0002] The traditional method of recommending programs is to obtain the type, album, and singer of the program that the user listens to, and recommend the corresponding program type, album to the user, or the program of the singer to the user. However, most current music recommendation methods are based on collaborative filtering. The method has the following disadvantages: the music recommendation method implemented by the collaborative filtering algorithm cannot reflect the user's preference changes in time, and the update of the program correlation is slow, and the program correlation is easily affected by the noise data. Contents of the invention [0003] The present invention aims to solve the above problems of the prior art. A TV program recommendation method with improved recommendation accurac...

AI Technical Summary

Problems solved by technology

[0002] The traditional method of recommending programs is to obtain the type, album, and singer of the program that the user listens to, and recommend the corresponding program type, album, or singer to the user. However, most current music recommendation methods are based on collaborative filtering. The method has the following disadvantages: the music recommendation method implemented by the collaborative filtering algorithm cannot reflect the user's preference changes in time, and the program correlation update is slow, and the program correlation is easily affected by noise data

Images