Construction of Music Multimedia Retrieval and Measurement Quality Model based on Ranking Learning Model

Abstract

Reasonable music feature library construction methods are of great significance to the realization of accurate music retrieval. In view of the limitations of current music multimedia retrieval, a ranking learning model is introduced in this paper, constructing corresponding models by fusing local and global indexes, constructing and mapping the attributes and characteristics of music multimedia using feature matrix to clean the data; looking on the global search, nodes are used to search, aiming to improve the efficiency of music multimedia retrieval and measurement quality model construction. The results show that the ranking learning model can effectively support the construction of music multimedia retrieval and measurement quality model.