Information Retrieval Based On Semantic Web About Areas Of Bronze

With the Semantic Web's emergence, a perspective combined artificial intelligence with the human world is proposed. While the concept of cultural fields as a reference model has been developed for many years, resulting in a number of low-level ontology for the field of culture. Chinese bronzes went through about 15 centuries, which was the cohesion of the Chinese nation's glorious history. In order to help experts better to research bronze information, this article applied the semantic web to bronze for the first time and eventually implements a system which can be conveniently used by cultural expert to query, digital, browse.This article has constructed information retrieval model from the perspective of the Semantic Web, by constructing domain ontology, the semantic information was added to the database, which was extracted form a lack of semantic data structures to extract knowledge and constructed knowledge base. The data was organized into machine understandable format, and then using Semantic knowledge analysis and retrieval on knowledge base. The main studies as following:(1) To build the ontology oriented to bronze field:based on characteristics of CIDOC CRM and the ABC model, combined with China's own characteristics of bronze, it would been eventually designed domain ontology for the bronze.(2) Based on seven layer architecture of the Semantic Web:using ontology metadata for semantic annotation for each bronze and building a knowledge base, then the inference engine of the knowledge base to explore and resolve, finally a prototypal framework for the Semantic Web of the Chinese bronze was build.(3) System implementation:ontology intelligent retrieval servers, Web servers, knowledge base, user interface, protocol analysis and other modules are include in the system. Through the introduction of Socket Communications, the retrieval server and web server were distributed deployment, which reduced the load on each machine and improved query speed. Multi-threading mechanism enhances the ability to throughputing and data rate of exchange on retrieval server. A multi-angle, multi-level search method were provied for experts according to their different requirements, which including fuzzy matching based on keywords, based on cultural information integrated query, query-based plastic artifacts and related inquiries. At the same time, in order to improve system's scalability, it was also provided for expert to add, delete, modify operation interface with knowledge base.According to the search results, this system achieved the desired goal, which was certainly improved in the recall rate and precision compared with the traditional retrieval method. It was provided a certain instructive for subsequent application in the field of bronze.