The Application And Research On Social Network Analysis In Evaluation Of Domestic Academic Journal

Academic journals plays an increasingly important role in the spread of information, and is an important source of information. According to statistics, Research workers of various field require documentation which 70% to 80% from the journal. Therefore, we need to give these massive source of information ( journal) for a scientific, rational, objective and fair assessment, find the information contained in density, high quality academic journals on behalf of disciplines.From the 20th century 30s, scholars began to study the evaluation of academic journals, and gain some far-reaching research achievements. The domain scholar begin to work in this area in the 20th century 70s, and make great achievements on the basis of the mature theory of foreign. But we also see the current journal evaluation methods, tools, indicators are insufficient. The journal evaluation results are very different.Social network analysis is a mature method of sociology, its use of "significant is the importance " basic principles, to discover the important influence on the network node is unique. Based on this viewpoint, we can use social network analysis to discover the significant influence node in the citation network, and these node should be very important in the network. We use this method to evaluate the library information science and physics journals, and reach the following conclusions:(1)InDegree have statistical relevance of the original evaluation results, can be directly used as an indicator of journal evaluation. (2)Closeness centrality lacks of distinction, and requires more strict, so unsuitable for journal evaluation. (3) Indegree Centrality, Outdegree Centrality, Degree Centrality and Betweeness Centrality can be used together to judge the status of the journals, such as find the knowledge resource, intermediary, recipient of knowledge, knowledge controller and so on. This information is an important reference for academic journals evaluation, and this information is the original evaluation methods can not provide. (4) Papers journal volume has a certain impact on the degree indexes, but the effect is not absolute. Therefore, we can use the average centrality as the original complement. (5) We can Use the social network visualization tool and clustering tool to provide a more intuitive reference for the classification of academic journals.