Document Recommendation System Based On Scientific Research Knowledge Base Design And Implementation

In recent years,as scientific research has continued to deepen and expand,more and more research-related documents have appeared on the Internet.The increase in the number of research documents has led researchers to find articles that they are interested in becoming slower and inefficient.This paper designs a scientific document recommendation system based on this point,according to the user's history browsing records,summarizes the user's browsing rules,and then recommends to users the scientific documents that they may be interested in.The main research contents of this thesis include the following aspects.The scientific research knowledge base mentioned in this paper is a database containing millions of scientific research documents.The first step is to crawl scientific research documents.In order to better enrich the research knowledge base,the system First,we need to crawl more documents from the Internet to ensure that the scientific research knowledge base is rich in documentation and meets the needs of use.In order to ensure the speed of crawling,a crawler with a distributed master-slave structure is used to increase the speed of crawling.And fixing the small domain name on the URL ensures that the crawler crawls scientific documents on the designated website.Then the analysis of scientific research documents and the acquisition of user data,because the crawled documents are not all in a unified format to meet the needs,so you need to use big data technology for cleaning before you can use the database.The document search function in the system is not This part of the system,but the realization of other parts of the entire system,the system can directly obtain the user's browsing operation data,after cleaning to obtain a unified structure of the data to use,then the feature extraction of the document,to the document Labels,so that they have clear attributes rather than long texts,and finally according to these tags based on the design of the scientific document recommendation algorithm,first according to the algorithm to the user to recommend the tag list.