Design and Implementation of Chinese Journal Data Analysis System

Abstract

Journal industry is developing rapidly and the amount of journal paper data continues to increase. Journals have become an important way for researchers to publish research results and consult literature. As far as the platform of China National Knowledge Infrastructure and database exchange service center is concerned, there are nearly 1800 kinds of core journals and more than 5000 papers are updated every day. How to conduct research and analysis on such huge journal paper data? In this paper, a Chinese Journal Data Analysis System is designed by using machine learning algorithm, apriori algorithm and time-series keyword co-occurrence method. This system is to perform statistical analysis, text similarity analysis, hot spot analysis and hot spot prediction on the abstract, authors, keywords, funds, publication date and, etc. of the China National Knowledge Infrastructure’s chinese core journal paper data. The experiments prove that the analysis system has high accuracy and good stability. Meanwhile, the design and implementation of the system provide guidance for the development of data analysis systems.