Approach for multi-dimensional associated heterogeneous engineering document semantic retrieval
YAO Yuan-gang,LIN Lan-fen,DONG Jin-xiang
DOI: https://doi.org/10.3785/j.issn.1008-973X.2011.02.012
2011-01-01
Abstract:Aimed at the problem of insufficiency of document semantic comprehension and association in current engineering information retrieval,according to the analysis of information retrieval characteristics in engineering environment,a heterogeneous engineering document information retrieval method was proposed,which included content analysis,semantic modeling,multi-dimensional association,semantic inference,expansion,and query process.Ontology based comprehension was utilized to capture engineering semantic and build semantic annotation base,which realized unified description of heterogeneous document content.Multi-dimensional association architecture was introduced focusing on documents.On the base of semantic comprehension of document content,ontology based internal association was expanded to multiple associations.The result shows that multi-dimensional association can associate documents in product life cycle heuristically,improve query navigation,and support semantic inference and retrieval expansion.