Based On The Interaction Intents Modeling Information Excavation Technology Research

Today,Web information retrieval technology has become an important way for people to query information,along with the expansion of the information,people have had to spend a lot of time to search the information through their need.Search engine has become the most common tools for people to retrieval information tools.The tool involves the information retrieval,database,data mining,artificial intelligence,distributed processing,natural language processing,and other fields of theory and technology,thus has comprehensive and challenging.But,at present most of the services provided by the search engine can't makes users satisfied,especially when the user's search intention unclear and needs constantly change over time.So-called the user search intention,we can understand by searching for the information or resources which the user wants to obtain,and can quantify hope to get the retrieval result set for the user.The combine of Interaction intents modeling and visualization of the user interface can provide feedback to guide the user through search intention for exploratory search,to help users find new information and improve their ability of information to explore.Therefore,the interaction under the intention of the search engine becomes an important research topic in the current,this paper main research work is as follows:(1)studied the web-based interactive intention search model.Using web retrieval model and search intention model framework system to implement the information search,and on the radar map visualization of present a system for users to retrieve the result set.System will be estimated by the intention of the visual rendering on the radar map,and presents a novel visual interface,in the radar map,Angle corresponding to the distance,similarity search intention from the center of the radial distance corresponds to the relevance of search intentions.This method can be observed through successive to update the online learning model,which can make immediately when receiving feedback subsequent forecasts.This paper will use a hybrid approach system of evaluate the advantages of interactive intention modeling.This article will deformation of two kinds of interface interaction intents modeling(radar map interface and intention list)were compared with the traditional information search system.The results show that the interaction intents modeling greatly improve the user's work efficiency as well as to the quality of information retrieval.(2)according to the traditional web text clustering algorithm was improved,through topic extraction,feature extraction,text clustering more than three steps to realize the theme of web text clustering,compared with traditional web text clustering algorithm,the proposed method fully considers the topic information of web text,the experimental results show that the theme of web text clustering,the proposed method of accuracy than text clustering method based on k-means better.