智能穿戴设备中的自动问答系统研究

Research on Automatic Question Answering System in Intelligent Wearable Device

如何使穿戴设备更加智能,对用户提出的问题快速准确地做出回答是该领域研究的焦点。针对此问题,提出了构建自动问答系统的方法。该问答系统通过穿戴设备接收用户问题,并在分词处理后运用贝叶斯模型算法进行分类,采用基于知网的句子相似度算法进行问题匹配,实现针对用户以自然语言描述的问题,自动从问题库的问答对集中抽取答案,并返回给用户。实验表明,该方法有效地提升了穿戴设备的智能程度,对用户提出问题的回答具有较高的准确率。

How to make the wearable device more intelligent and be able to answer the user's questions quickly and accurately is the core problem in this field.To solve this problem,the author puts forward a method of constructing automatic question answering system.This system receives user questions through a wearable device.After participle processing,Bayesian model algorithm is used to classify,and sentence similarity algorithm based on Hownet is used to calculate sentence similarity for problem matching.Then according to the questions described by the user through the natural language,the system automatically extracts the answers from the Q&A of the question base and provides the answers to the users.The experimental results show that this method can improve the intelligence of the wearable device effectively,and has a high response accuracy to the questions raised by the users.