Research On WEB News Content Classification Technology Based On Evolutionary Fuzzy Rules

In today's big data era,the network generates a large amount of industry data every minute,and the amount of data is almost beyond the capabilities of existing processing and analysis tools.In addition,with the development of modern technology and living standards,people's understanding of things is constantly evolving.Among them,Web news is a daily contact with people.As a very representative data,Web news content has always been acquired by people with its diverse content,real-time and constantly updated.Know the necessary channels for current events.In the face of explosive news content,how to quickly and directly find the news content that individuals want to know,and adapt to people's continuous improvement of knowledge,this is a hot research topic.So how to adapt to the current and trend of Web news category is complex and constantly updated and evolved,and the classification of evolutionary fuzzy mechanism for a large number of news data has very important research value.This topic focuses on the classification of Web news content under the evolutionary fuzzy rules,extracts from Web news content,and classifies news content based on evolutionary fuzzy rules.The research contents are as follows:1.Establishment of a news content classification model framework based on evolutionary fuzzy rules.Based on the understanding and experimental evaluation of various existing methods,and understanding how the existing framework is established,this paper constructs a Web news content classification model framework based on evolutionary fuzzy rules.Two important core parts of the framework were established: the identification and extraction of Web news content and the classification of news content based on evolutionary fuzzy rules.Transform the classification problem of text into the processing of natural language.2.Identification and extraction of Web news content.The recognition and extraction of Web news content is the premise and necessary condition of the classification method based on evolutionary fuzzy rules.Through the analysis of the construction classification model,it is determined that the collection and retrieval of Web news content is very important,which is related to the effect of the final classification.This paper extracts and recognizes the text content of Web news and the video subtitle content of news photos.The text content department proposes a topic content extraction method based on web page classification usage and web content segmentation.The web news image and video frame image subtitle content part proposes the use of edge clustering preliminary positioning,and then uses BP neural network and spectrum analysis.Non-text filtering positioning method.Experiments show that compared with some existing methods,the two methods are suitable for all aspects of the environment in the subject matter of our research,and the accuracy of the latter for text positioning methods is greatly improved under certain conditions.3.Classification of news content based on evolutionary fuzzy rules.Based on the obtained Web news text information and video text information,the pre-processing steps of the extracted text information are performed without affecting performance.In this paper,the combination of mutual information and word frequency-reverse document frequency(TF-IDF)algorithm is used to filter out the term set to improve the execution efficiency of the algorithm,and then improve the fuzzy classifier(FRB)based classifier to realize the Web.Evolutionary fuzzy classification of news content.Compared with some existing methods,the two methods have certain effects in the context of the subject matter of our research,and the accuracy of the latter for text positioning methods has been greatly improved under certain conditions.The outstanding contributions of this paper are: in the stage of recognition and extraction of Web news content texts,this paper proposes a non-text filtering and positioning method based on bp neural network and spectrum analysis;in the classification stage of Web news content,this topic combines mutual information and The word frequency-reverse document frequency method is used to filter out the terms with low description ability.By improving the existing fuzzy rule classifier,the classification of Web news content based on evolutionary fuzzy rules is realized,so that the accuracy and content of news content classification are improved.The rate has improved.