Research On Reasoning And Recommendation Of Cyberspace Security Knowledge Graph

The rise of the Internet not only promotes social development,but also brings risks and challenges to the security of cyberspace.For coping with this problem,the society urgently needs talents in the field of Cyberspace Security.However,due to the complexity of cyberspace security problems,it is difficult for ordinary developers and learners to learn and understand the knowledge of Cyberspace Security in a short period of time.As a solution with high flexibility and strong expressive ability,the knowledge graph helps to make the fragmented knowledge in Internet data produce actual value.However,there may be errors,incompleteness,and inconsistencies in the knowledge graph.Therefore,it is necessary to obtain new knowledge or detect wrong knowledge through knowledge reasoning.After having the knowledge graph,it needs to be combined with other tasks to maximize its effect.For the problems faced by this paper,constructing a recommendation system based on the knowledge graph is a good solution.Therefore,the focus of this article is to conduct knowledge reasoning and recommendation on the cyberspace security knowledge graph.The main contents are as follows.First of all,this paper artificially constructed the cyberspace security ontology and supplemented the relevant triplet data from the Internet open domain and the data from the Q?A community platform Stack Exchange to the ontology through the ontology matching method based on machine learning,and then used the graph database Neo4 j to store knowledge graph.Then,this paper realizes knowledge reasoning through an improved model of neural tensor network.Referring to trans series methods,this paper uses the relationship vector into the model directly.At the same time,the attention mechanism is used in the neural tensor network to capture the relationship between the triples and the reasoning process.Since there are a large number of parameters in this model,this paper also decomposes the parameter tensor slice in the model to reduce the number of parameters and prevent over fitting to a certain extent.Finally,this paper studies the recommendation algorithm based on the cyberspace security knowledge graph.While using the knowledge graph,this paper combines user and post data in Stack Exchange for knowledge recommendation,and adds user-defined auxiliary information that has often been ignored in the past.While recommending matching posts for users,the recommendation algorithm can also recommend related knowledge through the knowledge graph,which can provide convenience for developers and learners.