Research On The Description Method And Application Of Tacit Knowledge Of Research Achievements

In recent years,with the rapid development of science and technology in China,the number of scientific research results has exploded.In order to more efficiently obtain tacit knowledge from scientific research results and make targeted recommendations to scientific researchers,Chinese scientific research papers were selected as the main research objects from various types of scientific research results,and the following work was done:(1)Extract the subject words of scientific research papers.The subject terms of the paper are different from the keywords defined by the author.Use natural language processing technology to clean and normalize scientific research paper data,extract structured data in papers based on rules,extract topic words of papers based on TF-IDF algorithm,and obtain relevant label data,which is finally used to form a knowledge graph Entity in.(2)Complete the discovery and screening of new words in scientific research papers.Based on the word formation rules of Chinese characters and the N-gram algorithm,the sliding pane scan is performed to extract the collection of the thesis words.Since not all the obtained candidate new words are current new words,the candidate new words need to be processed.Calculate the boundary degree of freedom and internal solidification of the words,set a threshold to filter and screen the words that do not meet the target new words,and finally get the target new word set.(3)Discover the relationship between words in the paper.The research has obtained the set of target new words,constructs the word vector matrix to calculate the similarity between words,and then uses manual to screen the upper and lower vocabulary relations,so as to realize the extraction of the hierarchical relationship between words in a semi-automatic way;using association analysis The method explores and discovers the non-hierarchical relationship between the technical points of the paper.(4)The construction and application recommendation of knowledge graph.Based on the NEO4 J graph database,a map of scientific research paper achievements is constructed,various entities in scientific research papers,as well as the attribute value types of entities and the relationship types between different types of entities are determined,and the knowledge map of scientific research papers is displayed visually.Based on the knowledge graph of scientific research papers,collaborative and social recommendation algorithms are used to make targeted knowledge recommendations to scientific researchers.