基于Louvian社区发现算法的科研成果挖掘技术研究

Research on scientific research achievements mining technology based on louvian community discovery algorithm

本文对科研成果数据库中涉及应用科学、社会科学、自然科学、生命科学与生物医学、艺术与人文五大学科的部分论文数据进行统计分析,发现每年论文数量增幅最大的是应用科学,而计算机科学的研究热度最高。以此为基础,基于Neo4j图数据库绘制高频词汇语义网络图,进一步明确了实体关系,使社区内部的关联更为紧密。采用社区发现算法开展数据挖掘,总结成功社区及其科研团队的特质。同时,基于PageRank算法制定论文数量和质量“双高”人才名单作为各学科科研合作的人才储备。

This paper makes a statistical analysis on the data of some papers in the database of scientific research achievements involving five university subjects:Applied Science,social science,natural science,life science and biomedicine,art and humanities.It is found that applied science has the largest increase in the number of papers each year,while computer science has the highest research popularity.On this basis,based on the neo4j graph database,the semantic network diagram of high-frequency words is drawn to further clarify the entity relationship and make the association within the community closer.This paper uses community discovery algorithm to carry out data mining,and summarizes the characteristics of successful communities and their research teams.At the same time,based on PageRank algorithm,the"double high"talent list of paper quantity and quality is formulated as the talent reserve of scientific research cooperation in various disciplines.