Citation Semantic Link Network Community Discovery Research

The citation networks are a kind of network that are composed by the citing reference, the cited reference and the citation relationship between the citing reference and the cited reference. The citation network is a knowledge network. Through the citation network we can understand the process of knowledge generation and dissemination. However, while the citation network provides people with a lot of knowledge, the citation network makes people fell into a swamp and sometimes confused. When one search documents by keyword, Usually find a great deal of relevant literature which makes people difficult to gain literature they want fast and accurately. If the original citation network has a semantic representation capability, it will provide a more efficient way for query retrieval and citation network studies.The citation network lacks the semantics and can not be reasoned. To research the semantic representation of the citation networks and semantic reasoning we need that. In order to make the citation network express semantic information and possess reasoning ability, here in this paper we first combine the semantic link network and citation networks, and then propose a citation semantic link network model that make the original citation network not only posses the semantic representation ability but also has the reasoning abilities. We can obtain the implicit citation semantic link and the semantic relationship between the citations according to the rules of citation semantic link. Our research could be helpful for the sort of development, the core work, and the overall knowledge structure.With the rapid increase of knowledge, citation network has become a large-scale complex networks. In the citation networks, each literature has played different roles for each of the flow of knowledge. Some literatures are at the core of the flow of knowledge and have played a pivotal role. These literatures have a significant impact on the field, while the impact of the others are small. In order to find the semantic community in these research areas and important documents in the semantic community, we proposes a community discovery algorithm of the citation semantic link network. With this we can find the semantic communities of the different semantic field in the citation network. We can get some information about a specific paper. Such as its position in the network and the role in the process of knowledge flows. We can get the big picture of knowledge evolution and help improve the efficiency of macroscopic management and policy decision.Based on above ideas, we research community discovery of citation network through a combination of semantic Web, semantic link network, citation networks and complex network theory. The main research work and innovations in the following areas:(1) We propose a citation semantic link network model for the representation and reasoning of the semantic relationship between the citations. In order to make the citation network express the semantic information and has reasoning ability for the full and efficient access to citation information, we combine the semantic link network and the citation networks, and propose the citation semantic link network model. The original citation network will not only have a semantic and with reasoning ability, but also can obtain the implied citation semantic link and the new semantic relationship between the citations according to the citation semantic link network reasoning rule. The Citation semantic link network model provides a high efficient way for the retrieval of citation networks and the research of complex networks.(2) We propose the citation semantic link network community discovery algorithm based on the similarity calculation. This paper firstly gives the semantic similarity calculation between the citations, and then according to citation semantic link network community discovery algorithm, we can find the semantic communities of citation semantic link network and important documents related to research areas. This are of great significance to reveal the dynamic structure of science, understand the issue forward in all areas, and predict the future direction of development and hot spots.(3) We use Matlab to construct experimental system, simulate the citation semantic link network community discovery process, analysis the experimental data. The proposed citation semantic link network community discovery algorithm is shown to be feasible and effective.