GENERATION OF SPECIFIC SOLVERS FOR QUERY-ANSWERING PROBLEMS WITH SKOLEM FUNCTIONS

摘要

In this research, we propose a novel technology for solving the Query-Answering Problems (QA Problems) including Skolem functions by generating a specific solver corresponding to each QA Problem in the Semantic Web. A specific solver is constructed by using the knowledge (conditions) analyzed from a given QA Problem and is only used to deal with this given problem, since we think by this way the computation efficiency can be much higher than generating a general solver for all QA Problems based on the common knowledge. We first get the clause set including the Skolem functions from a given QA Problem based on the theory of meaning-preserving Skolemization. Then, a specific solver corresponding to this given QA Problem will be generated by using the clause set based on the Bottom-up solution. Moreover, we have also developed the technology to suppress the size of the generated specific solver for reducing the usage of memory in order to deal with larger scale QA Problems. In the final part of this paper, experimental results will be used to prove the efficiency of solving the QA Problems by generating specific solvers.