Robot knowledge graph node updating method based on voiceprint recognition

Abstract

The invention belongs to the field of service robots and artificial intelligence, and particularly relates to a robot knowledge graph node updating method based on voiceprint recognition. The method mainly comprises the following steps: starting voiceprint recognition, collecting voiceprint information of a target person, querying a local or cloud database, judging whether the voiceprint information is in a knowledge graph database or not, if not, adding a new figure entity, taking the voiceprint information as an identification tag, actively asking a target person to obtain basic informationof the target person so as to add a new triple instance, reasoning the relationship between the target person entity and other entities, distinguishing different character entities, and carrying out personalized man-machine interaction selection in a targeted mode. According to the method, character entities are distinguished through voiceprint recognition, so that a robot determines different entities in conversation with the robot, entity information is actively added into a personalized knowledge graph of the robot, new triples are continuously inferred in interaction, the relationship between the entities is updated, and the knowledge graph is perfected.

Description

technical field [0001] The invention belongs to the field of service robots and artificial intelligence, in particular to a method for updating nodes in a robot knowledge map based on voiceprint recognition. Background technique [0002] With the continuous deepening and development of robot technology, robots are becoming more and more intelligent, and robot knowledge graph is the key technology for its intelligent progress. At present, most service robots only rely on fixed knowledge graphs for interaction, and do not have knowledge graph update methods. There are two ways to update the knowledge graph nodes of some advanced robots: one is to manually update the nodes, and the user needs to manually upload the relationship between entities and store them in the knowledge graph to achieve the purpose of updating nodes; the other can realize automatic update Nodes, relying on natural language processing techniques, understand entities appearing in sentences in question-answ...

AI Technical Summary

Problems solved by technology

[0003] Among them, the method of manually updating nodes is too limited and cumbersome. Users or developers need to enter triples by themselves. The method for the machine to infer triples to update the knowledge map is very flexible. The robot will continuously improve its knowledge map in the process of communicating with people, but it has great limitations in learning the relationship between characters and entities. The robot It is impossible to know the character entity communicating with him, so it is difficult to collect triples related to the character entity in the interaction of natural language

Images