Research On Dynamic Monitoring And Analysis System Of Patent Information Based On Ontology

In the knowledge-economical time, knowledge is a kind of key resource propelling economical developent. Patents, as a important representation form of knowledge, are not only the important reference information sources for research scholars to know novel technology, enlighten thinking and avoid the repeated research and development, as well as improve the research and development efficiency, but also the essential intelligence analysis sources for managers and decision-making personnels to analyse the trend of industry and international technology, and efficiently collocate some invisible resources, as well as constitute strategic planning etc.At present, the patent information gathering and analysis technologies researched and developed by some scholars and institutions at home and abroad have many problems, such as most of patent information gathering technologies are mainly based on keywords and their weights, and don’t consider semantic relations among words, synonymy, polysemy etc problems, and have resulted in the lower precision and recall of patent information retrieval, as a result, the patent analysis results based on the retrievaled patent information aren’t too precise, and can’t reflect the real status in the patent domain; the patent information gathering mode is still staying on human-aided, and can’t adapt to users’ dynamic and changeful requirements of real time gathering and analysis of patent information; the patent information analysis indexes and functions provided by many patent information gathering and analysis systems aren’t too comprehensive, and the visual ability of analysis results is low, so which can’t satisfy users’ fexible and changeful individuation and interaction analysis requirement of patent information.For solving the aforesaid problems, this paper takes patent information gathering and analysis as research objects, and designs and develops a dynamic patent information monitoring and analysis system based on ontology, which can not only carry out a semantic expansion to users’ gathering expression by using ontology, thus improving the comprehensiveness of patent information gathering, but also real-time and dynamically monitor and gather the Deep Web patent information on the Internet, and carry out a visual analysis to the collected patent information with different indexes, consequently, solving users’ dynamic, changeful and real time individuation and interaction gathering and analysis requirements of patent information. System’s basic implementation processes are as follows: Firstly, we carry a word segmentation, stopwords filtering etc preprocessing operations to users’ retrieval sentences by using chinese morphology analysis system (ICTCLAS) developed by Chinese Academy of Science, thus forming users’ origin gathering expression, and with our constructed patent domain ontology carrying out a semantic expansion to user’s origin gathering expression, forming the extended semantic gathering expression. Then according to the extended gathering expression, we compile gathering programs to real-time and dynamically gather free patent data from the patent data sources, and after removing the repeated data and purifying data, we save them into the patent database. Finally, we utilize the visual software Open Flash Chart (OFC) to implement a visual show of patent information analysis results, and take costume domain as example to verify the feasibility and practicability of systems.