Intelligent sensing human-computer interaction system based on temporal knowledge graph
Under the background of the improvement of social informatization,the demand for human-computer interaction tech-nology is also increasing.In order to improve the efficiency and user experience of human-computer interaction system and expand the application scenarios of human-computer interaction system,knowledge representation learning Trans series model is introduced to optimize the key technologies on the basis of temporal knowledge graph.Through experimental analysis,it can be seen that the fit de-gree of the research algorithm in GraphQ dataset is the highest,reaching 98.2％,in which the unsupervised learning method has un-derfit.The accuracy of word meaning understanding,syntax analysis and context understanding under the research method is above 95％.The average accuracy of the research system is 97.6％,which is 25.3％and 31.6％higher than that of the system based on re-current neural network and unsupervised learning,respectively.The optimal path length of the robot in the research system is 153.4m when executing multi-objective dialogue requests.To sum up,the human-computer interactive dialogue system studied has high fit-ting degree and high accuracy.
temporal knowledge mapintelligent perceptionhuman-computer interactiondialogue system
万方数据
维普
