| In recent years,with the rapid development of artificial intelligence technology,research institutions at home and abroad began to study how to let the computer to solve problems automatically.The existing automatic problem-solving system mainly obtains problem-solving knowledge from problem texts and related resources by manual or semi-automatic way,and the difficulty lies in the acquisition of implicit problem-solving knowledge.Based on the theory of three-way concept analysis,this paper automatically obtains the implicit problem-solving knowledge from the problem texts through the construction of decision-making formal context,the construction of three-way concept lattice and the extraction of decision rules,so as to improve the problem-solving effect of the existing automatic problem-solving system.The main work of this paper is as follows.(1)Semantic analysis of problem textsAiming at the problem of acquisition of key feature information in geographic text,this paper proposes a method to acquire the entity association classes and their attributes based on domain dictionary,which provides the semantic support for formal decision context construction,three-way concept lattice construction and decision rule extraction.(2)Extraction of three-way concept lattice rules from problem textsAiming at the acquisition of implicit problem-solving knowledge in geography texts,a method of extracting three-way concept lattice rules based on object-induced and attribute-induced three-way concept lattice is proposed,and the obtained rules are optimized by calculating confidence and support.(3)Design and implementation of automatic problem-solving prototype systemIn order to apply the acquired implicit problem-solving knowledge to the actual automatic problem-solving task,this paper designs and implements an automatic problem-solving prototype system by adding three-way concept lattice rules into the ontology knowledge base. |
PDF Full Text Request