Study on the Keywords Indexing System Based on Linked Triple Technology for Chinese Scientific and Technical Literature

摘要

For Chinese scientific and technical literature, keywords indexing and relevance construction is one of the most important links in literature processing. Based on the ontologies that registered in the ontology management platform and the use of web services techniques, the triples of concepts and relation from the knowledge base are assigned for the keywords recognition and indexing for scientific and technical literature. With domain dictionary the system can recognize the domain terminology and with word segmentation and grouping technology to recognize the unlisted words. Build links between the triples from ontologies and the indexing information of the keywords and form conceptual relation networks that deep inside the literature. The indexing results can support the deep analysis and mining work of literature texts. According to the corpus test, the system can carry out literature indexing and words linking at a relatively high speed of 86 pieces/s, with recall of 65%, precision of 69% and comprehensive performance of about 66%.