Research Of Internet Financial Search Engine Based On Semantic Context

Search engine is one of the most commonly used tool for information inquiry.As the improvement of requirements for the accuracy of searching results,the semantic gap between requirements and searching results of traditional search engine based on keywords matching and sorting algorithm become more obvious.Besides,it cannot understand the internal relationship between queries and web content.In order to make it better,the paper proposed a search engine model based on semantic context.Firstly,we use topic model to build the mapping set between unstructured document and potential topic vector,combining with words weighting model to further optimized the inverted index space.After that,we extend the semantic context based on corpus with filtered keywords by document generation model.In addition,we explore the searching method of traditional search engine further and search the context with it.At last,we calculate the relevancy of the consequent documents based on weighted semantic context with ranking models,return to users after sorting.The experiments are based on the internet financial news data,comparing with traditional search engines in aspects of indexing space,search result,and performance.Experimental verifications have demonstrated,our search engine based on semantic context save the indexing space and can provide more accurate search services.