Research And Implementation Of Semantic Search System In Printer Operation Based On Ontology

With development of science and technology, the functions of electronic device are getting more and more, and the devices are becoming more and more popular in people's daily life. Users may encounter with a lot of problems when using the device. Device manufacturers and users pay more attention to how to provide users with service to help them effectively solve operation problems.Currently, web-based search can provide a convenient and efficient service, users can apply the vendor's online system or search engine (such as Google, Baidu) to query the solutions of the problems. But these search are based on text matching, cannot understand the meaning of queries, the results are not satisfactory.In this paper, the research aim is to realize an ontology-based semantic search system in printer operation domain, which can help user apply daily used non-professional words to query the domain specialized knowledge. Because of the deficiency of traditional information retrieval technology, the paper introduces the ontology technology, utilizes ontology to describe the domain knowledge and analysis user query at the same time, to achieve the semantic level query rather then text-matching based search. The paper mainly employs the semantic web technology, such as ontology, RDF and SPARQL to realize the semantic query. Firstly, according to analyzes the printer operation knowledge, constructs the query ontology and printer ontology. Secondly, applies query ontology and printer ontology to carry on semantic analysis of keywords, uses the device ontology to represent the device domain knowledge in formal of RDF triples, analyses the user query keywords with query ontology, then through the rule which defined in the paper, convert the keyword-based query to SPARQL semantic query. Finally, adopts the traditional search engine ranking model to rank the results.Based on 16 printer user manual, paper constructs query ontology (defined 944 concepts) and printer ontology (defined 1642 concepts,3129 RDF triple), a prototype has been implemented, According to the experiment, the result shows that the system can achieve semantic query, the prototype's accuracy is 84.4%, has met the system design requirement.