Design And Implementation Of Vertical Patent Search Engine

Copyright,patent and trademark right are three main contents of intellectual property right.Among the above three contents,patent plays a significant role in an enterprise's development.The influence and development prospect of an enterprise are determined by the quality and quantity of patents it has.However,considering priority and the large amount of patents in the world,we have to guarantee the originality of our discovery before applying for a patent.The main purpose of this article is to introduce the development process of a vertical patent search engine written in JAVA.This system is based on APACHE's open source projects---LUCENE and SOLR,along with Pao Ding Jie Niu word segmentation method.Several expansions were made by analyzing the original functions and actual needs of LUCENE and SOLR,which improved the original searching function and added the characters of patent at the same time.The system contains three main functions which are data retrieval,keyword highlighting and packet statistics.It proves that the system can finish patent retrieval at TB level with high speed and accuracy through system testing.In present thesis,background knowledge and motivations are firstly discussed.The author then clarifies some definitions of related technical terminologies,and discusses user requirements and technical bottlenecks.In the subsequent section,the design of current searching system is broken down into three aspects-structure design,process design,and module design,and each are elaborated in details.The code implementation of the system is achieved and presented.The test of the system performance shows satisfactory results.In the last section,advantages and disadvantages of current work are summarized,and the scope of future work is discussed.