Research On Semantic Information Retrieval Model Based On Concept Lattice

Semantic information retrieval in the early1980s began to discuss, but due to thelimitation of the traditional retrieval technology, semantic information processing hasnot developed very well. Accompanied by retrieval technology upgrading, theimprovement of ability to deal with natural language and the development of artificialintelligence, many famous domestic scholars began to research some semanticinformation retrieval model. After further research on ontology, we found ontologyhas a big advantage to reveal the semantic concept and to describe aspects abouthandling of semantic concept. Based on the semantic level, applying ontology got agreat development in the field of semantic information retrieval.Concept lattice is put forward by a professor Wille at Germany in1987,including the intent and extent of concept: intent defines the semantic of concept;extent defines concrete things corresponding the semantic concept. It can dig deeperlatent semantic, the hierarchical structure of formal context can describe concepts,classes, the relationship contained between attributes. Hasse diagram can visualize theconcepts got by matching. So concept lattice is one of the powerful tools for dataanalysis and data organization, it is widely used in many fields, the main aspects ofapplication in the field of information science contain such as information retrieval,knowledge organization in the digital library, knowledge representation, ontology anddata mining analysis, etc.The design idea of this paper is: Semantic information retrieval systems candistinguish the mapping relationship between documents and key words by semanticrelations between the concepts, in order to improve the precision of informationretrieval. Revealing the semantic relations between the concepts needs to rely onformal context. Formal context is a triple consisted of the intent(attributes),extent(objects) and the relationship between concepts. Objects-attributes in the formalcontext is corresponding to document-key words. Concept lattice can expound theintent and extent of the concepts through the hierarchical strucure of formal contexts and mapping relationship about object–attribute; can dig deeper semanticrelationship. We design a concept retrieval mechanism and construct the semanticinformation retrieval model based on concept lattice by using ontology modelinglanguage.In particular, according to the top-down modeling method this paper describesthe framework about semantic information retrieval model based on concept latticeconsists of four layers, respectively is: human-computer interaction interface layer,service layer, data storage layer and network layer (operating system and networktransmission layer), which focuses on service layer and data storage layer constitutingthe core of semantic information retrieval model; on the basis of analysis describesbriefly the relationship between the four layers; establishes three principles ofbuilding semantic information retrieval model based on concept lattice on the basis ofa detailed analysis: concept lattice based on ontology, following the thoughts of usingconcept lattice in retrieval on the macro and secondary organization of retrievalresults. The concept lattice based on ontology expounds from six angles: conceptdescription, the relations between concepts, knowledge mining, knowledgeorganization, knowledge reasoning and knowledge retrieve in this paper. Followingthese principles and the microscopic structure model we constructs semanticinformation retrieval model briefly, and constructs concrete semantic informationretrieval model by expanding the core mechanism of brief model.In view of the design model, we select ball games on the sports meetings as theresearch objects for case analysis, and compare the two aspects of recall ratio andprecision between retrieval results and keywords retrieval results, and conclude that itis significant to apply concept lattice in the field of semantic information retrieval inthe end.