Design And Implementation Of Patent Retrieval System Based On Solr Technology

With the continuous development of science and technology,the amount of patent data continues to increase,the patent data management becomes more and more difficult,and it is increasingly difficult for enterprises to quickly and effectively extract valuable patent information.In response to the above problems,this paper designed and implemented a patent information retrieval system based on Solr.The main work includes the following aspects:First of all,because it's complex to realize technically the patent information retrieval system,the Solr full-text search engine technology and the IK Analyzer tokenizer plug-in were used to realize the retrieval function of the system.Solr implements the inverted index function by encapsulating Lucence.By using IK Analyzer's customized tokenizer plug-in,terms or phrases can be searched so as to easily look up patent documents.Solr search technology is mature and stable,it supports a variety of data formats,and better supports the diversity of patent data formats.Secondly,the system was optimized by hardware and system architecture to ensure the retrieval speed of the system.On the hardware side,the server memory and CPU configuration were upgraded.The traditional hard disk was replaced with an SSD hard disk to improve the index reading and writing speed.In terms of system architecture,the Field,the Solr's base unit,was merged and weighted to improve the retrieval speed.Thirdly,a distributed search scheme based on SolrCloud and ZooKeeper was adopted.The Solr index was deployed as a cluster to ensure the high availability of the retrieval system.SolrCloud and ZooKeeper distributed deployment can make it access other Solr data when a single Solr service is unavailable,without affecting the overall search function.Finally,this paper completed the basic construction of the patent information retrieval system with Solr,and improved the retrieval speed through the upgrade of the server and the optimization of the Solr's architecture.At the same time,the distribution of SolrCloud and ZooKeeper and cluster deployment of Solr for this system ensured the high availability of the retrieval system.Performance indicators and function of the retrieval system were tested,and the test results showed that the function and performance were in line with the expected results.