Patent Information Analysis And Applied Research, Based On Data Mining Technology

Due to the publication of patent documents, the public will stand to gain from the technical knowledge of inventions because of the published information in these documents. Therefore, patent information is a very useful resource of technical knowledge for researchers, engineers and scientists who are involved in the work of companies, research institutes and governments. Moreover, data mining id believed to be the trend of the techniques used to exploit patent information. As the most form of patent information id text, text clustering which is one of the most important data mining operations has very high applied values.Data mining technology is an effective approach to resolve the problem of abundant data and scanty information. It is the research frontier within the information science field currently. The related researches and applications have greatly improved the ability for decision supporting.It has been deemed to a field that has broad prospect of application in database research. Data mining has been widely used in many fields and has resolved many commercial issues, made a good effect on our society.Because patent has both exterior characters which should be analysed by quantitative analysis, and internal characters which should be analysed by qualitative analysis, so data mining would take much value in patent information analysis. Association analysis method is getting useful information from these massive, inaccurate, noisy data. Since Agrawal et al. started to research the relationship among item sets based on commercial database, there are more and more scholars have done more deep studies about this topic. Using the cluster analysis method to analyse text, we need to carry on the pretreatment of the text. Firstly, we need to pick up words from text, then finding out the most important words and translating them to data that can be accepted by computers. After that we can use those methods of clustering to cluster text. The level cluster method decompose one big database into several smaller ones one by one until some kind of condition had been satisfied. Comparing with another common method of cluster analysis which called k-means, level-clustering algorithm has been used in uncertain environment that cannot know the number of clustering.In respect to the application of data mining, aim to the actual demands of mining patent information, this paper analysis the application and uses of the patent documents, and research the patented technology and the special features of the patent documents, and study the wealth of technical information, legal information and business information which is contained in patent documents. Then try to find the approach to mine patent documents with data mining technology. Finally, the paper takes hybrid electrical vehicles field of patent information platfom of TOYOTA JIDOSHA KABUSHIKI KAISHA as the main application background, with the data mining methods and theories, and associate the results of cluster with association rule mining association data mine in patent documents, and proposals some suggestions in further mining patent information.