Research On Patent Information Acquisition System Based On Multi-Agent System

In order to study and the establish of Beijing intellectual property index system, public information service platform and intellectual property early-warning mechanisms, to improve early warning capabilities and high-tech industrial competitiveness,a large number of accurate patent is need urgently as analysis and study basis. Currently most of the government and enterprises acquire patent information artificially or through a search engine to search and download. With low efficiency and accuracy, traditional patent acquisition process is not conducive to analysis.In this paper, the process of acquisition patent information from United States Patent and Trademark Office (USPTO) web database which includes patent searching, interface analysis, patent download and information extraction has been studied. The methods such as Deep Web technology, Multi-Agents technology and information extraction technology based on XML have been used in patent acquisition. The purpose of the paper is acquiring patent information accurately, rapidly and automatically from USPTO database.In this paper, a patent acquisition system has been designed based on Multi-Agents Systems, which includes user interface subsystem, central control subsystem, patent web pages acquisition subsystem, patent basic information extraction subsystem and referenced patent information extraction subsystem. The cores of each subsystem were user interface agent, central control agent, spider control agent, extraction control agent and referenced control agent. The paper discusses the structure and algorithm of communication and scheduling of the agents. Based on Java language and JADE environment, a prototype of patent acquisition system has been implemented and the experimental results show that the prototype system has high performance and efficiency.Major contributions of the thesis are as follows: firstly, various technology and methods were used such as MAS, information extraction based on XML and some open source tools in order to acquire patent information from USPTO web site.Secondly, the core algorithms of patent acquisition system have been studied and designed which include central control agent scheduling algorithm, spider control agent scheduling algorithm, extraction control agent scheduling algorithm, referenced control agent scheduling algorithm, the process of patent web pages acquisition, the process of patent information extraction, web pages cleaning and filtering algorithm and the rules of information extraction. Thirdly, a prototype of patent acquisition system has been implemented which can download, extract and store patent automatically from USPTO web database with high efficiency.