A Natural Language Processing Method of Chinese Instruction for Multi-legged Manipulating Robot

In this paper, a natural language processing (NLP) framework is proposed to build the Chinese natural language interaction between human and the multi-legged manipulating robot which has both locomotion and manipulation functions. The Chinese natural language instruction is transformed into formal representations by this method. Firstly, Cascaded Conditional Random Fields (CCRFs) is employed to analyze the syntax of the instruction. Letter-based features are used in each layer of CCRFs to solve problem caused by word segmentation. In order to understand the modification relationship between entity classes in chunk, one judgment method based on Support Vector Machine (SVM) is proposed. To solve the problem of the large number of core verbs in instructions generated by the multiple robot motion types, a classification of verbs based on Naive Bayes classifier was presented. And the semantic framework of each type of verbs is established to determine the necessary and unnecessary roles for each kind of verbs. At last, several experiments are carried out and the results of each step of framework are presented to demonstrate the effectiveness of the method. It is instructive to understand Chinese natural language instructions for robots with complex motion.