基于包装产业大数据知识图谱的智能问答系统设计

摘要

传统智能问答系统能够进行简单的问题答复,但对于问句信息比较相似的问题不能准确判别.基于此设计基于包装产业大数据知识图谱的智能问答系统.通过知识图谱问句映射进行问句信息框架设计,建立大数据知识图谱数据库,为问句信息的判别提供稳定的数据环境;将问句信息转化为eFAQ的判别语句,进行问句信息的理解;利用相似度计算判别相似问句信息的相似度,实现近似问句信息的判别问答.实验数据表明,设计的智能问答系统能够对相似问句信息进行精准的判别,并实现智能问答.%The traditional intelligent question-answering(Q-A)system can answer simple questions,but it cannot accurate-ly identify the questions that have similar question information. An intelligent Q-A system based on big data knowledge map of packaging industry is designed. The question information framework is designed by using question mapping of knowledge map. The database of big data knowledge map is established to provide a stable data environment for question information identifica-tion. The question information is transformed into the eFAQ decision statements for question information understanding. The infor-mation similarity of similar questions is identified by means of similarity calculation to realize discriminating answering of similar question information. The experimental data shows that the designed intelligent Q-A system can accurately identify similar ques-tion information and realize intelligent question and answering.