Analysis And Design Of Competitive Intelligence System Of Enterprises Based On Text Mining

With the rapid development of Internet,more and more enterprises get the information they want on the Internet and extract competitive intelligence from this information.In terms of collecting intelligence information,general search engines can collect simple information,but it is difficult to conform to the enterprise personalized information and deal with the large number of enterprise information resources.It becomes difficult for the enterprises to classify information processing and analysis the intelligence information.On the one hand,the explosive growth of information makes it facilitative to the business of intelligence gathering.On the other hand,it also makes enterprises face more severe challenges.How to obtain accurate and reliable information from the mass of Internet information resources becomes a required course for the companies.Enterprises urgently need a competitive intelligence system to collect intelligence,mining and analysis of intelligence.The main work is as follows: after deeply investigating the enterprise intelligence demands and analysis the enterprise competitive intelligence system detailed needs,the system is divided into competitive intelligence gathering,competitive intelligence analysis,competitive intelligence services of three subsystems.Through analysis the features of each subsystem and use case information,it determines the needs of the system.In this paper,first is to get the page information with a search engine.The search engine searches and specifies the initial URL(Uniform Resource Locator)by specifying the keyword web crawler URL information,with the two ways to download a web page and obtain information page.These preliminary results are the download pages of semi-structured and there are many redundant knowledge,so the preliminary results drawn using text preprocessing techniques screened to remove noise data,further structured competitive intelligence text,and then to Chinese text segmentation and removal of stop words and other operations,so that the structure of the text is easy to use text mining.In text mining,the use of two methods: text categorization and text clustering,the user can advance a given topic and let the system automatically learn.Finally the text is divided text set.The text classification is based on SVM(Support Vector Machine)to achieve a category specified topic category text.In the text clustering,it uses the K-means method without specifying subject categories,and the documentation set automatically divided into several similar clusters,which realizes unsupervised machine learning methods,effective text information organization,summary and navigation.By using text mining technology,users only need to match the search category labels,pay attention to their desired category cluster,finally users can quickly find the information they need.Through the establishment of enterprise competitive intelligence system,the enterprises achieve the collection,processing and analysis of competitive intelligence.It is convenient for enterprises to find information by text mining processing intelligence information to facilitate business intelligence and improve the use of intelligence information in a text value.The system helps corporate management to get a potential meaningful competitive information,which helps business managers to make the right decisions on the current situation and improve work efficiency of the enterprise.It has high practical application value.