The Design And Implementation Of Text Summarization System For Economic News

With the advent of the Internet era,the information on the network is growing explosively.The Internet has become an important way for people to obtain information.People can't read all the texts to obtain important information in limited time.Text summarization technology emerges as the times require.At the same time,the economy develops rapidly,more and more people begin to pay attention to the development of the economic field of enterprises,and the economic news websites and WeChat official accounts emerge one after another.However,on one hand,the texts on some popular economic news websites or WeChat official account is too long,requiring readers to spend a lot of time reading them.On the other hand,most designers choose to manually extract the important information from texts,which will cost much energy and time.Therefore,text summarization technology can be applied to economic news.On the one hand,this way is friendly for readers.On the other hand,it provides design ideas for economic news websites and WeChat official accounts.First of all,this paper introduces the text summarization technology,selects some classic text summarization algorithms to introduce,and compares their results on the self-made data set.Then,I choose the TextRank algorithm as the basis.Next,in view of the text characteristics of economic news,I combine a reward and punishment mechanism with the TextRank algorithm,put forward an improved single text summarization algorithm based on TextRank algorithm.And then I designed and implemented a text summarization system for economic news.This system also uses crawler technology,automatic news classification technology based on keywords in news headlines,key information selecting technology for different types of economic news,keyword extraction technology,multi-text summarization technology,sentence similarity calculation technology and text similarity calculation technology.The system architecture uses the Flask framework,the front-end design uses the Bootstrap framework,and this system uses the MySQL database to store data.This paper also introduces the design of some important modules in the system in detail,including news information acquisition and processing module,multi text summary module and news collection module.At present,the system has passed the functional and non-functional tests,runs in a good state,and can be extended and maintained regularly.When users use the system,they can browse the key information of recent hot news,browse the key information of news according to the news category,add the interested news information to their favorites for future reference,and the system can also generate summarization and other information for the text uploaded by users themselves.The system can also meet the users' needs in response time,data accuracy and other non-functional indicators.