Research And Application Of Semantic Retrieval Systems Based On OWL

With fast development of information society, The information on Internet increases very fast, and begins to show great complexity and multiformity. Now that traditional information retrieval(IR) technic based on keywords can no longer satisfy people's needs for information searching. The main problem is that the traditional IR system can't understand information resources' inherent meaning and the relations between them, that is to say, lacking the ability to understand semantic information. The information people want to find out is usually related to structure of domain knowledge, but the results turned back by keywords matching only contains the literal information, which is far from the information people really want. Semantic retrieval technic is the most promising method to solve this problem, which has become one of the most important issues for reseach in IR.This paper mainly researches how to apply ontology to IR systems, to implement semantic retrieval. The paper starts from the status quo of traditional IR technic, to analyse its main problems, then adequately expatiates how semantic information retrieval solves these problems. The paper then introduces the key technics of semantic information retrieval, such as the concept of ontology technic, ontology languages of RDF/RDFS and OWL, Jena frame for semantic web application, semantic inference and so on. Based on these technics the paper makes a deep research on the working process and mechanism of semantic retrieval, and makes the design of semantic information retrieval systems.At last this paper implements a meeting-paper semantic information retrieval demo system. Then the paper shows the feature of semantic information retrieval and the effect of semantic inference mechanism within semantic information retrieval. Base on this practice, the paper summarizes the demo system's deficiency, makes a prospect for the development of semantic information retrieval.This paper's main new ideas are that it designs a IR entry interface with strong expressing ability, which could use ontology's semantic feature to enhance users' ability to express their retrieval needs and does a deep research on how to enhance semantic information retrieval systems' intelligence by semantic inference.