Subject-Oriented Analysis Of Patent System

Patent literature is important form of science and technology achievements. Itcan quickly and comprehensive show the latest achievements in science andtechnology. It can give guiding significance to our enterprises, scientific researchinstitutions and government. Therefore, scientific research institutions at home andabroad pay more attention to patent literature search and patent information analysis.They hope dig the valuable knowledge through the patent data. It can help enterpriseor countries understand the industry dynamic, technology development trend andtechnological opportunities. So they can improve the accuracy of decision-making andthe ability of early warning.In recent years, some scholars at home and abroad have researched in patentinformation extraction, collection, analysis. But they have no further research in thepatent information processing, patent analysis indicators, patent multidimensionalanalysis, data mining and the subject analysis. Although some experts put forwardsome patent analysis indicators, patent analysis model, but these analyses needmanual analysis and consume a lot of time on patent analysis. It will greatly increasescientific research personnel’s workload.We design and implement subject-oriented analysis of patent system according topatent analysis demand and data source. First of all, we extract patent data into localpatent database according to the patent analysis demand. Because of storage format,download method, system factors and human factors, local patent database has manysome "dirty data". It needs the appropriate design of ETL(Extraction-Transformation-Loading) for cleaning patent data sources. And then weget a clean foundation patent data which can provide high-quality patent data forsubject-oriented patent analysis.According to the different theme analysis model, weestablish some subject-oriented data marts, multi-dimensional data cube and keyperformance indicator through the extraction and integration data. Use this system wecan analyze patent data from the different themes. At the same time this system canmake patent map according to analysis result. Finally this system can publish theanalysis result to information platform in order to realizing cross-enterprise sharingand centralized services.The innovation of this paper lies in: first, this paper make use of SSIS (SQLServer2005Integration Services) technology to achieve foundation data warehouse.At the same time, it designs theme model library and KPI library, and using SSIS toimplement theme data mart, establish dimension table and fact table structure;Second,it make use of SASS (SQL Server2005Analysis Services) technology to achievemultidimensional Analysis model and theme data mart slice, parts and KPI analysis.Last, it choices Excel2007as a result display tools for providing the basis for business decisions.Using this system, users can efficiently integrate patent information andautomatically depict the evolution of a technological history. At the same time, thissystem can automatically measure innovation capacity of enterprises by patents. It canput forward a new prediction and provide a more efficient decision.