Frontier Topic Identification In Scientific Research

With the development of a new generation of high and new technology based on computers,scientific research has developed from a high degree of differentiation to a cross-comprehensive development.Library and Information Science is a new subject that combines the subject of library and information.Its professional learning involves the courses of library science,information science and archives among the three.It not only contains "information",but also "management",which can be used as an important reference for research on interdisciplinary.The discipline of library information and archives management in our country is in a stage of alternating between tradition and emerging.Research hotspots and research topics are constantly changing.How to accurately mine and identify potential research topics and current research hot topics has become a research problem that needs to be solved urgently.This article combines the advantages and disadvantages of Implicit Dirichlet Allocation(LDA)topic mining algorithm and social network analysis algorithm.A topic recognition algorithm based on the combination of LDA topic mining and social network is proposed.From the macro level,the overall research field is divided into small areas.In various fields,we are conducting micro-excavation of the links between the subject terms.In order to give representative names of each field and core subject terms under that field.This paper uses LDA topic mining algorithm,co-word analysis algorithm and social network analysis algorithm to divide the core topics in the field of library and information into five.They are the topic of information service;the topic of intelligence analysis;the topic of digital resources;the topic of knowledge management;the topic of library research.This article uses reasonable text mining methods to conduct related research to eliminate subjective wishes.It is processed,classified and predicted by the computer itself.Combining multi-source data for analysis and research effectively solves the limitation of the unity of scientific research data.This paper proposes a topic recognition algorithm to make the research topic more clear.At the same time,it also reduces the professional threshold of subject hotspot research,allowing more new researchers to join the discussion of subject hotspots.At the same time,it also lowers the professional threshold for subject hotspot research,allowing more new researchers to join the discussion of subject hotspots,and enhances the academic confidence of new scholars.The subject mining research provides practical and effective reference opinions for the development of library and information field.The research results of this paper have certain application prospects and practical value.