The Research Of Semantic Problem Solution System Based On Domain Knowledege

International Collegiate Programming Contest is the world wide programming contest hold by Association for the Computing Machinery (ACM). In order to get good achievement, each school will train good students to attend it. With the development of Internet technology, there are many ACM problem solutions in the current Web space. But many students can not really make full use of them.There are mainly three reasons for this phenomenon: 1) The current Web is just a container of "Information", of which the contents are only suitable for students, not with machine-readable semantics, which causes the lack of enough semantics for machines searching the Web. 2) The current Web describes programming solutions unilaterally, which leads that students could not get full information and solve the problem thoroughly. 3) The solution in the current Web has no relation with each other, which leads that students could not realize the difference among the solution and machine-readable semantics information has no structure.Aimed to solve such problems, this paper creates programming solution semantics system based on domain knowledge and design many programming problem and their solution as examples to develop researches. Machine-readable semantics solution with some certain relationship and helping students solve problems are the biggest difference between this paper and the current Web.Firstly, this paper analyses the domain knowledge model by ontology and design the programming solution ontological model. This model meets the concept model which is the reaction of the programming solution in people head. It is suitable for ACM programming and each solution in ACM programming can use it. Though this model, programming solution will be semantics. It will be useful to combine all the solutions.Secondly, this paper designs the structure of programming solution semantics system based on ACM domain, which combine all the semantics solutions. The concept lattice on programming solution is built by Format Concept Analysis. The concept lattice can describe the relationship among semantics programming solutions. And this paper designs the concept of programming solution similarity by using Format Concept Analysis and Network flow minimum cost maximum flow. The concept of programming solution similarity is used to describe similarity between two programming solution. Though concept lattice and programming solution similarity, students can realize all the solved problems and unsolved problems thoroughly from the whole to the local.At last, this paper describes one application of programming solution semantics system based on the above theory. It shows the correctness of the research work in this paper.