基于约束的自然语言问题到OWL的语义映射方法研究

OWL已经成为重要的知识表示和交换方式.OWL驱动的问答系统是比搜索引擎更具挑战性的研究.本文致力于解决为自然语言问题获取正确的OWL解释.考虑到问题理解的复杂性,影响因素的多样性,本文提出一种基于约束的语义映射方法.该方法分解问题成为变量,索引OWL知识成为变量域,抽象约束形成函数.并建立了目标函数完成问题变量在候选OWL知识库中获取合理解释的过程.该方法基于网络智能研究院知识库和问题集进行了初步评估.实验结果表明:和传统基于关键字匹配的算法比较,本文提出的方法使可解释问题比例增加了18%左右.

A Constraints-based Semantic Mapping Method from Natural Language Questions to OWL

The goal of an on-line ontology-based question-answering system is to automatically derive answers from ontology knowledge bases without demanding additional information or intervention from users.This paper focuses on solving mapping task from question variable to OWL elements,which belongs to component of question understanding in question answering system.The paper presents a principled approach,which builds on tokens of question and syntax or semantic relations from NLP into set of variables and functions of constrain,and a sound optimization-based assigning mechanism to find sound substitute in OWL knowledge for question variables.A preliminary evaluation on the International WIC Institute knowledge and questions is presented.