The Modeling And Analysis Of Knowledge Network Based On The Community Structure

"Knowledge is the ladder of the progress of mankind". Every since the human society, there was knowledge. While realized its value, we began to research, that the research of knowledge never interrupted. At present, the world economy has been changed from the dominant industrial economy to knowledge economy, knowledge of the study also appear particularly important.Study from the perspective of information, we can consider that knowledge of human civilization in its own process of natural, social and other areas of the human spirit of understanding and conclusion, which is a manifestation of information. So we should first take into account how did these specific"information"came out. Previous studies, many scholars have made quite a lot of research achievements, and also put forward many knowledge representation methods, such as concept map, the semantic web, mind map, but these methods have their respective advantages and disadvantages, which has the suitable study theories and applications area. Based on the complex network theory, we develop a new kind of knowledge representation (knowledge network) ground on the complex network. In Knowledge network, we focus on the performance of the link between knowledge and knowledge of a specific knowledge in the whole system (that is, the entire knowledge networks) and the role of the position. The node of the knowledge usually represent a particular subject area of general knowledge points (concepts, definitions, etc.); edge between knowledge for all kinds of semantic association, such as component relationship, inheriting relationship, extension relationship; etc. The subject area of knowledge has formed a complex network of knowledge network characteristics.As a relatively complete theoretical system, a complex network of various complex systems can be modeled and analyzed, which affect the overall (to) the local characteristics of network is quantitatively described. Knowledge system is a complex system, in any discipline knowledge, concepts, and each significant between disciplines, the same concept of hierarchical relationships of the concept of knowledge, the same subject such contact with each other, and compared different areas of knowledge is sparse contact. These phenomena appeared that complex network theory to the study of knowledge have a greater applicability. Based on the knowledge Network established by the nature of the physical meaning and mathematical properties of the in-depth study and found that actually a number of "groups" consisting of each group of nodes within a relatively close link between, the connections between groups relatively sparse. Those keep pace with"community structure"in complex network. Then this paper will study the modeling of the knowledge network and analysis the characteristic, which is based on the "structure" basic theories.Through the research, the knowledge network has obvious characteristics of "small world", especially, the local groups and the overall sparsest. That is to say, a domain knowledge which we study, that must relatively smaller in the other domain knowledge, which we can call"community structure". In order to master the whole knowledge, we can research the "community structure" first. In other words, a smaller group of knowledge can show the characteristics of all knowledge fields. That also provides theory basis for the teachers and the students.In this paper, the specific jobs were as follows:The first part: using the complex network theory, all the concepts, definitions and the relationship between them of "database theory" in computer software and theory were summarized. Then coding, modeling and input database based on the complex network according to their relationship. The basic concept and definition of modeling method: the keywords (similar to the teaching of the knowledge points) as the network node, if two key nodes appear in a concept, you can connect one side of the two.The second part: after the analysis of the knowledge of network structure and characteristics of network, we can analysis each node (knowledge point) the relationship between the discussion of the complex nature of network statistics, such as the average path length, clustering coefficient, degree distribution, betweenness centrality, coreness and so on, in the knowledge of the specific embodiment of the network.The third part: through the research and analysis of network knowledge are all connected graph club, so we put forward a kind of this complex web searching algorithm of community. Calculate the number of community networks, the number of the nodes in every community and each node respectively belonged to. This study of the characteristics of the entire network has significance in understanding the practical value of the network structure.In this paper, the research based on the specific knowledge system and the relationship between them has unique characteristics. The database system introduction of computer theory was selected as the research object, which was used widely. The steps of the research as following: First, according to the complex network theory, we put forward for the modeling mechanism of network database knowledge combined with the characteristics of knowledge database, and then through the simulation, found it face the actual situation perfectly. Second, we proposed for the characteristics of the knowledge of web search algorithm, which lays the foundation for better study. Third, we abstracted the knowledge point and the relationship from the database to "dot" and "edge", analysis the network using the theory of complex network modeling, which lays a foundation for study computer knowledge modeling.The number and size of community structure in complex network has great influence on the overall characteristics. In order to improve the algorithm efficiency, a new community structure searching algorithm was put forward, in fact, it achieved the good result.