| Information technology acts as a very important role in the knowledge management.The application of information technology in data information has become a hot issue in the current research.The disorderly data can be shown in the knowledge using this technology.Journal Submission is an important channel for research results published.Because of the type and number of journals are more and more,it is very difficult to grasp the Potential knowledge in the literature using the current manual operations,and sometimes get the wrong conclusions.Based on this premise,this paper researchs the key technologies of Growing Hierarchical Self-Organizing Map(GHSOM) in the text cluster analysis,and combined with bibliometrics and information diffusion theory, there are 1401 thesis from metal mining magazines and conference proceedings which are carried out text cluster analysis,and try to reveal the current metal ore mining projects focus on research directions and trends.In this paper,the Growing Hierarchical Self-Organizing Map is the tool for information visualization,it revealed the research and development trends in academic. Through the text data of the publication of academic papers as input,taken the terms of a major theme as the characteristics,then produce the self-organizing map which behalf the research theme using the term relations.Finally,the thesis would be projected onto the the self-organizing map.According the time of papers publication,we can confirm the focus of research and development at different times,understand the change of theme.This paper study the domestic metal mining in the area,the main research contents include:1.Generate vector space model literature procedures using TF/IDF algorithm.2.GHSOM neural network literature ore cluster analysis based on the MATLAB.3.Propose the analysis method of the literature clustering results combining of bibliometrics and information diffusion theory.4.Reveal the current metal ore mining projects focus on research directions and trends through the establishment of cluster analysis model to analysis 1401 literatures.In this research,we look forward to speed up a lot of complicated analysis of the process of Periodical Literature,and to combined research of metal mining areas with information management technologies to analysis the text message and to be quantified, in order to provide the reliable guidance for predict the development trends of some research field.And hope that through the use of this technology in the direction of other subjects to help identify research trends and and improve research management efficiency and scientific management.In addition,network technology,the future of search engines through the use of technology in this study to establish the theme of knowledge map to search for classified information and accordance with the hierarchy will be shown in a very bright future. |
