基于知识图谱的齿轮传动智能问答系统

鉴于市面上搜索引擎的搜索结果繁杂且针对专业领域的问答结果很少的情况,提出了一种基于优化后Att-BiLSTM-CRF深度学习模型的问答系统的构建方法.将机械专业的问答数据与智能问答技术结合起来,实现了导入问答文档进行自然语言处理后自动生成对应齿轮传动知识图谱的功能.在用户输入问题后,系统会通过文本相似度算法和Viter...

Automatic Question Answering System of Gear Transmission Based on Knowledge Graph

At present,the search results of search engines on the market are complex,and there are few questions and answers for professional fields.In view of this situation,a question answering system construction method based on deep learning and optimized Att-BiLSTM-CRF model is proposed.Taking gear transmission as an example,this system combines the question answering data of mechanical specialty with deep learning and intelligent question answering technology.Finally,it can automatically generate the corresponding knowledge graph after importing the question and answer documents processed by Natural Language Processing.When user inputs problem,the text similarity algorithm and Viterbi optimal path algorithm are used to match keywords.Experiments show that most of the answers are reasonable and accurate,which can be widely used in online and offline intelligent teaching of engineering colleges and intelligent explanation of related exhibitions.