Text recommendation method and device, server and medium

Abstract

The embodiment of the invention discloses a text recommendation method and device, a server and a medium, and the method comprises the steps: obtaining N first keywords, and determining an occurrencefrequency sequence of each first keyword, wherein the occurrence frequency sequence of any first keyword is obtained by combining the occurrence frequencies of any first keyword in the dynamic text set at multiple moments; obtaining a reference frequency sequence, and determining M second keywords in the N first keywords according to the sequence similarity between the reference frequency sequenceand the occurrence frequency sequence of each first keyword; determining a unit text set corresponding to the M second keywords in the target text set; clustering the plurality of texts to obtain a plurality of text clusters, and selecting a target text cluster from the plurality of text clusters, the number of target texts in the target text cluster being greater than a number threshold; and determining a to-be-recommended text according to the text theme of the target text. Therefore, the text recommendation efficiency and accuracy can be improved.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular to a text recommendation method, device, server and medium. Background technique [0002] With the rapid development of electronic technology and Internet technology, there are more and more multimedia data, and the recommendation system has also been developed rapidly. The recommendation system can extract the data needed by users or current social hotspot data from massive multimedia data. At present, most recommendation systems use the method of manual recommendation, that is, it is necessary to manually select the data needed by users or current social hotspot data from massive data, which is inefficient and difficult to accurately select the data to be recommended. And then affect the recommendation effect of the recommendation system. It can be seen that how to improve the recommendation accuracy of the recommendation system has become a current research hotsp...

AI Technical Summary

Problems solved by technology

At present, most recommendation systems use the method of manual recommendation, that is, it is necessary to manually select the data needed by users or current social hotspot data from massive data, which is inefficient and difficult to accurately select the data to be recommended. And then affect the recommendation effect of the recommender system

Images