News event mining method and device, computer equipment and storage medium

Abstract

The invention discloses a news event mining method and device, computer equipment and a storage medium, wherein the method comprises the steps of: acquiring a plurality of news texts; for each news text, according to a semantic vector of each vocabulary in the news text, obtaining a semantic vector of the news text, wherein the semantic vector of each vocabulary carries semantic information of corresponding contexts; according to a similarity among the semantic vectors of the plurality of news texts, carrying out clustering on the plurality of news texts to obtain a plurality of clusters; according to release time distribution of the news texts contained in the plurality of clusters, determining a target cluster from the plurality of clusters; and according to a subject of the news texts contained in the target cluster, determining required news events. By the method, contents of the news texts can be effectively understood and shown in a form of the semantic vectors, and by carrying the semantic information of the corresponding contexts in the semantic vectors of the vocabularies and considering correlation of the contexts among the vocabularies, mining accuracy can be improved.

Description

technical field [0001] The present invention relates to the technical field of natural language processing, in particular to a news event mining method, device, computer equipment and storage medium. Background technique [0002] Text mining can summarize, classify, cluster, and correlate the content of document collections, etc. Mining news texts is conducive to automatic retrieval of news content to meet the needs of network users to retrieve news information. [0003] The traditional news text mining method is based on the Bag Of Word (BOW) model. After data preprocessing such as Chinese word segmentation and stop word removal are performed on the news text, the content of the preprocessed news text is analyzed using the BOW model. Then use the clustering algorithm to cluster, and get the required news events according to the clustering results. [0004] However, because the same word may have different meanings in different contexts, sentences containing the same word h...

AI Technical Summary

Problems solved by technology

[0004] However, because the same word may have different meanings in different contexts, sentences containing the same word have different meanings due to different word orders, and the BOW model is based on the number of different words contained in the text and the presence of each word in the text. The number of occurrences in the text is expressed as a vector, without considering the contextual relationship between the words in the text, so the mining accuracy of the news text mining method based on the BOW model is low

Images