Research On Construction Of The Integrated Vocabulary Based On Interoperability Technique Of Indexing Languages

It is very difficult to retrieve network information nowadays because there are various subject headings, thesaurus, classifications and taxonomies and the same concept is usually expressed using different subject terms or class number among different websites and systems. It is an ideal mode for users that one query can get the results from many databases, which primarily needs to realize interoperability among different indexing languages. In recent years, the information professionals have been working on the issue of the interoperability among different indexing languages and have presented many methods to resolve it, including automatic matching and switching, intermediate lexicon, integrated vocabulary, mapping, translation etc. Based on these studies, they also have accomplished many projects which have brought convenience for users to search information.Through studies of interoperability technique between indexing languages, the paper construct a compatible system, which regards Classified Chinese Thesaurus(CCT) as core and can be extended constantly. It can realize interoperability between Chinese Library Classification(CLC) and library classifications from domestic and abroad, between Chinese Thesaurus(CT) and special domain thesauri, controlled languages and natural languages. Taking education class as a test, the paper selects several thesaurus, and library classifications to construct the integrated vocabulary. The data source comes from CLC, Classification for CAS Library, Dewey Decimal Classification(DDC), Educational Thesaurus, Social Science Retrieval Thesaurus etc.The paper mostly studies interoperability between different indexing languages. The methods of co-occurrence mapping and classes similarity algorithms are used to realize the interoperability between different classifications and some new modification are presented aiming at resolving the shortcomings of various algorithms. Such methods as automatic matching based on thesaurus's structure and mapping based on synonymy are used to realize the interoperability between different thesaurus. The paper also studies the switch from natural language to controlled language which provides access to natural language for users thus facilitating their retrieval and searching. The integrated vocabulary is constructed through mapping methods mentioned above and stored by compatibility matrixes including two forms: alphabetical compatibility matrix and classification compatibility matrix. In order to browse and use the compatible data, the paper adopts stand-alone mode, XML and ontology to display, and then provides to users various services.