Analysis Technique Based On Knowledge Of Subject-oriented Query

Data warehouse (DW) technique is a jumped-up field. Data warehouse which is separated from operational database in the enterprise,provide a subject-oriented,integrated data store and manager mechanism,and provide a strong platform for query and analysis of uniform historical data and integrated data. It supports information processing and decision making. However,the success of data warehouse greatly depends on front-end tool. Nowadays,the front-end tools of data warehouse mainly include query tools,OLAP tools and DM tools. These tools mostly suppose the users who have subject-related domain knowledge. So common users maybe have some difficulty to use them.In this paper,a framework of Knowledge-Based and Subject-Oriented Intelligent Query and Analysis system of DW (KB_SOIQA) is provided to improve the intelligent degree of DW front-end tool and the energy of query and analysis. This system combined the technique of DW,AI and DM. It provides a knowledge-based intelligent query mechanism of business concept terms and intelligent query answering technique. It can support knowledge-based ad-hoc query and provide more related information and knowledge for users. In this paper,main research work include:1) Research and design a Manager Tool based on Task-Chain Hierarchical Knowledge Base. Domain expert and knowledge engineer can use this tool to descript the background knowledge of Analysis subject domain. Based on database technique,knowledge base system provides a mechanism of centralized store and maintenance. System provides a visual tool to support the Knowledge Management.2) Research on domain-knowledge based intelligent query technique. Based on this intelligent query mechanism,users can not only query the data object which store in DW,but also query the business concept term which maybe not store in DW directly.3) Based on domain knowledge and metadata of DW,and combined data mining technique,Intelligent query answering technique is researched. Based on analysis of users' query,it provides more related information.