基于Python智能机器人多渠道知识库推送仿真

Python-Based Intelligent Robot Multi-Channel Knowledge Base Push Simulation

机器人知识推送是智能化发展的必然产物,当前相关研究成果存在召回率和推送结果用户满意度较低的问题,提出基于Python的智能机器人多渠道知识库推送方法。利用离线和在线的方式对访问和浏览信息进行识别,离线信息识别中,对采集到的信息结构进行分析,检测出信息的特征,动态添加字符串,将得到的特征与关键词知识库中的数据特征进行配准,判断出是否识别关键词;在线信息识别中,基于Python语言,分别结合百度云识别和云聊天以及百度云语音三个体系,实现信息的在线识别。利用信息语义相似度给出知识库推送的详细过程,对义项之间的相似程度进行计算,获取关键字或关键词的相似程度,将相似度比设定阈值大的信息保存起来,将此类信息推送给使用者。实验结果表明,上述方法查全率和用户满意度均较高,是一种可行性很强的知识库推送方法。

The knowledge push of robot is an inevitable outcome of intelligent development.Currently,the push results have the problems of low recall rate and low user satisfaction.In this article,a push method of multi-channel knowledge base Python was put forward.The offline method and online method were used to identify the access information and browsing information.In the offline information identification,the collected information structure was analyzed and the information features were detected.The character strings were added dynamically.After that,the obtained features were registered with the data features in knowledge base of keywords,so as to judge whether it was able to recognize keywords.In online information recognition,Python language was combined with three systems of Baidu cloud recognition,cloud chat and Baidu cloud voice to achieve the information online recognition.In addition,the semantic similarity of information was used to give the detailed process of knowledge base push.The similarity degree between the meaning items was calculated to find the similarity degree of keywords.Finally,the information whose similarity was larger than the set threshold was pushed to the user.Simulation results show that the proposed method has high detection rate,user satisfaction,and high feasibility.