Inversion Model Of Original Science Complex Network

In order to reduce tracking cost for advanced science,this paper studied the constraint inversion model of original science complex network.This mode changes the way of directly obtaining information to indirect reasoning.Firstly,the method try to track the international advanced technology skills,analysis the producing,manufacturing and selling process in the environment of big data.Then it will extract those technology and social attributes of the core product.And those units,institutions,companies that are directly related to the core node will be all found out.The association network of core product will be expanded based on collecting various channels of information.Finally,a complex network will be constructed and it will use for constraint inversion reasoning.In this paper,the main technologies of the constrained inversion reasoning include following contents:(1)The method of entity node vectorization is proposed in the process of complex network construction.The node vector element attribute is composed of technical attribute and social attribute.This can easily solves the problem which the social entity is difficult to formalize as a node;the score matrix is proposed to solve the problem which is difficult to quantify the strength of the relationship between different nodes in the network.Based on the node vector and the quantify weight,we can finally create an undirected complex network map.(2)A variety of constrained inversion searching algorithms are proposed based on complex network.And the combination of three strategies has higher efficiency and higher coincidence degree among all these searching algorithms.Through the classification of variable trigger events,we can greatly improve the speed of constraint inversion reasoning and the efficiency of node pruning.During the searching process,some node sets,path sets and constraint sets will be created.Among these sets,the node set is the foundation of network creating and expending;the path set is the basis of path searching and strategies executing;the constraint set is needed while searching and pruning in the whole network.(3)In this paper,coincidence degree is proposed to be a evaluation criterion of searching.and the percentage of same nodes between current path and target path will be defined as coverage degree.This paper also puts forward a path optimization evaluation criterion,which plays an important role in guiding the direction of network searching and deducing.During the process of constraint reasoning,correct rate will be pay much more attention than the efficiency of inversion.In summary,this paper realizes the research of constrained inversion of complex network.In order to show the rightness of the model proposed above,an experiment is completed in this paper and the result is compared to ant colony algorithm and genetic algorithm.The final results show that the theory proposed in this paper can be used to quickly inversion reasoning the source of advanced science and technology and it also has a good credibility.