Research Of The Model Of Enterprise Competitive Intelligence Collection System Based On Cross-Language Information Retrieval

Today, more and more attentions have been paid to Competitive Intelligence (CI) all over the world, especially in the enterprises. If the enterprises want to succeed in the fierce market competition, they must obtain the intelligence with the method of the collection of information. So, collecting really valuable information has become a necessary and important work for each enterprise. Enterprise which want to participating foreign competition, not only know their own circumstance about the enterprise themselves, but also know the circumstance about the rival. However, during to language barriers, users can not collect the relevant information in different languages. Therefore, enterprises urgently need an Enterprise Competitive Intelligence Collection System which can easily and quickly collect information in different languages. For this reason, we establish a model of Enterprise Competitive Intelligence Collection System based on Cross-language Information Retrieval, in the hope of solving language barriers in information collection.The main research content in this paper is elaborated as following:firstly, we review the research of competitive intelligence,competitive intelligence system,the collection of competitive intelligence and cross-language information retrieval,and introduce related problems and solutions of Chinese-English Cross-Language Information Retrieval. Secondly, the ontology was introduced in Cross-Language Information Retrieval, and introduce it's function in it. We present a model of Cross-Language Information Retrieval based on ontology and describe the functions and work flow in detail. This paper adopts the method of combining Bilingual Ontology and Bilingual Dictionary to solve the question of how to realize query translation and expansion:mapping the query to the concept in the Bilingual Ontology to realize translation, and expanding the query through the correlation between concepts; if the query dose not have concept in the ontology, then translated by Bilingual Dictionary and expanded by Thesaurus Dictionary. In order to improve the algorithm of concept similarity to make the query expansion fully expanded, this paper offers a way to compute concept similarity through integrating concept correlation of concepts in documentation sets and concept similarity in layer structure of Ontology. Finally, an Enterprise Competitive Intelligence Collection System model was established in this paper, which is based on Cross Language Information Retrieval, and detailed analysis was carried out.