Multi-ontology In Information Retrieval And Intelligent Human-computer Interaction Research

In recent years, the volume of information has been growing exponentially. People find it extremely hard to get what they need. How to organize, manage and maintain the huge information efficiently, and make them provide available service for users becomes a hotspot to research. Information retrieval is focus on how to search out the documents subset which is relevant to the users' query. An information retrieve system is composed of two parts in the whole: information organization and information retrieval. The improvement of the efficiency of the system needs a good organization as while as the satisfaction of the users needs a friendly Human-Computer interface.The traditional methods of information constructs index according to the external attributions, such as title, author and searching number. The defects of this method are that it cannot reflect the semantic attribution and it can only be searched exactly. Ontology as a modeling tool which can describe the concepts of the information system in the level of semantic and knowledge, can be used in information retrieval efficiently, and provide the IR (Information Retrieval) system with semantic backup. As the IR system becomes more popular, the traditional Human-Computer interaction method which depends on the pointers and keywords input cannot satisfy general user. The research on Human-Computer interface which adapts to non-professional users also becomes a hotpot, in which, NLP (Natural Language Processing) plays an important role.In this paper, the information organization in IR system and the Human-Computer interaction during retrieval process are researched in depth. By find out a sufficient way to organize information, cognitive help can be providing to the users and the quality of result can be improved. The main work of the paper is that, we research the related work of the research on ontology and multi-ontology and find out the shortcomings of the traditional IR models. The bases, an organization method using multi-ontology, an algorithm to compute the similarity of documents and an algorithm to broaden the concepts in multi-ontology are given. In the aspect of Human-Computer interaction, a method of questioning the users in natural language...