Family-oriented Service Robots Instruction Depth Information Recognition

With the rapid development of society and science,intelligent service robot has realized from the screen to People's Daily life,In order to realize intelligent service robot to serve people more conveniently and quickly,there are still many problems to be solved.Natural language instruction of shallow information have already completed the recognition task,This paper is to extract key information in the instruction on the indepth research,In this paper,a support vector machine algorithm is proposed to extract the hidden deep information in the instructions to help the robot to understand the speech instructions.First of all,this paper collects voice instructions from people to intelligent service robots in the home environment.The collected corpus is processed by word segmentation,part of speech tagging and other text processing.Secondly,the entity characteristics of the instruction statements in the corpus are analyzed,the conditional random field model is learned and the feature template for entity recognition is constructed,and the conditional random field algorithm is invoked for key information recognition.Based on the support vector machine model,LibSVM is used to extract the relationship between entities and store the extracted information and the relationship between entities.In this way,the execution efficiency of the family service robot serving people can be improved.Finally,because the XML knowledge base is easy to operate and fast,this paper stores the shallow information extracted from the instruction and the hidden deep information into the XML knowledge base.The experimental results and analysis of entity recognition are presented in this paper.This paper realizes the recognition system of extracting deep information of instructions.Figure 11;Table 11;Reference 41...