The Research On The Principle And Implementation Of Semantic Search Based On Latent Semantic Analysis

Nowadays Internet has entered a new stage of multivariate information,network information search technology has been widely used.The traditional information search system is generally based on the external keyword content matching,it is difficult to truly express the user's latent query request.Semantic search is a kind of information search method which is different from the traditional search method based on keyword matching,it analyzes the user's search intention accurately in the semantic level,and returns the semantic information which reflects the user's search motivation.Therefore,it is an urgent problem to be solved in the field of information search.Latent semantic analysis technology is a kind of information retrieval algebra model for knowledge acquisition and display,and is widely used in text classification and clustering,information extraction,information search,image retrieval and other fields.Especially in the field of information search,latent semantic analysis technology can express and store words and documents in the form of representing their semantics in a certain degree,and it can be used to search the information.The information search system can thus "understand" the user's query request to some extent,and provide the information needed by the user efficiently.On the basis of analyzing the shortcomings of the traditional search methods based on keyword matching and other semantic search methods,this paper deeply studied the principle of semantic search based on latent semantic analysis,and discussed the way to implement semantic search,used some information processing technologies such as the text content cleaning,Chinese word segmentation,keywords extraction,latent semantic analysis,semantic query expansion.Besides,we developed a semantic search prototype system based on latent semantic analysis,it could thus understand and deal with the user's query request in the semantic level,and mine the user's real needs and improve the precision ratio and recall ratio of information search system.In addition,this paper enriched and developed the mechanism which combines latent semantic analysis and semantic search,and it realized the successful transformation from the traditional search pattern based on grammar to the new search pattern based on semantics,it could therefore mine the available information resources efficiently and timely.As a result,it provided users with reasonable recommendation strategy.Finally,we conducted the training experiment and testing experiment on the prototype system by constructing the data set,and the experimental results showed that latent semantic analysis technology can extract the semantic relations between words in a certain degree,and it had a better performance in extracting large quantities of texts' latent semantic information and mining information that users really need.Therefore,that the research of latent semantic analysis and semantic search and the application of latent semantic analysis in the field of semantic search are of great significance to improve the search efficiency of traditional search pattern and perfect the theoretical system of information retrieval.