Hot topic finding method and system and hot topic finding equipment

Abstract

The invention discloses a hot topic finding method. The method comprises the steps that web page data is obtained, and segmentation treatment is conducted on the web page data to obtain a treatment result; characteristic extraction is conducted on the treatment result to obtain characteristic factors; the characteristic factors are inputted to an LDA model to obtain the probability corresponding to each topic in the web page data; the topics with the probabilities larger than a preset value are regarded as hot topics. Therefore, according to the hot topic finding method, by conducting treatment and characteristic extraction on the web page data to obtain the characteristic factors, extracting the topics in the web page data and the probabilities corresponding to the topics through the LDAmodel and regarding the topics with the probabilities larger than the preset value as the hot topics, the efficiency of finding the hot topics is improved. The invention further discloses a hot topicfinding system, hot topic finding equipment and a computer-readable storage medium. The system, the equipment and the computer-readable storage medium can also achieve the technical effects.

Description

technical field [0001] The present invention relates to the technical field of text processing, and more specifically, relates to a hot topic discovery method and system, a hot topic discovery device, and a computer-readable storage medium. Background technique [0002] With the rapid development and popularization of the Internet, various social networking platforms have become important channels for netizens to obtain and release information. It has become a very popular trend to freely express opinions on certain social hot issues or certain policies and express one's emotions through the Internet. This has enabled public opinion to be widely disseminated through the communication channel of the Internet. [0003] The hot topic discovery method in the prior art mainly adopts the clustering method, whether the selection of the initial clustering center is reasonable or not directly affects whether the subsequent clustering is accurate, and the selection of the number of c...

AI Technical Summary

Problems solved by technology

[0003] The hot topic discovery method in the prior art mainly uses the clustering method, whether the selection of the initial clustering center is reasonable or not directly affects the accuracy of the subsequent clustering, and the selection of the number of clustering centers also has a certain impact on the clustering results

Images