Topic Detection And Construction Of Technical Efficiency Matrix

Patents are an important source of technology information. It reflects the most active innovative information and cutting edge technology of the current corresponding field. Patent information mining of traditional collates the patent publication date, the applicant for a patent, nationality and other information. It based on computer visual display to assist the research on Patent Information. For recognition and extraction of deep information of patents, mainly rely on artificial reading. However, with the scientific research institutions and researchers of enterprise pay more and more attention to the patent, the patent number is exponentially increasing. The traditional method depending on the manual sorting for mining patent deep information not only waste a lot of manpower and material resources, but also is difficult to ensure the correctness. Therefore, this paper mainly studies how to mining patent deep information semi automated by computer.Through the studying related knowledge of text mining and information extraction, its application in the patent information extraction. Through the construction of patent technical efficiency matrix implements deep excavation of patent technology theme. This paper puts forward the constructing system of technical efficiency matrix to realize the semiautomation of Chinese patent technology efficiency matrix construction. Among them, the technical efficiency matrix system is mainly divided into 3 parts, the function words extraction system, the technical words extraction system, the construction of technical efficiency matrix. Function words extraction system extracts function words from the patent invention of improvement. It use Natural Language Processing and Machine Learning technology to realize the recognition and extraction of function words. The technical words extraction system recognizes and extracts the technical words in the patent abstract and patent name by text mining technology. Construction of the technical efficiency matrix processes technical words and function words and constructs the technical efficiency matrix which can be applied. Finally, this paper analysis on the technical efficiency matrix to implement the deep excavation of patent document. Proved by experiments, the scheme of Chinese patent technical efficiency matrix semi automated construction can greatly reduce the waste of human resources, and ensure the accuracy of results.