Multi-layer emotion calculating method based on mood index

Abstract

The invention provides a multi-layer emotion calculating method based on a mood index. The method comprises the steps of calculating the mood index at a current moment according to input stimulation; calculating the emotional intensity at the current moment according to the input stimulation and the mood index; calculating an emotional probability after state transition according to the emotional intensity and a state transition probability matrix, and outputting the emotional state with a maximum probability value. The method defines the mood index, quantifies a relation among mood, characters and emotion again, effectively fuses the influences of external stimulation, mood states, personality characteristics, emotion decay and emotion interactions on emotional change, can reflect the characteristics of human emotional change well and realizes personalized calculation for the emotional state of a robot.

Description

technical field [0001] The invention relates to a multi-layer emotional calculation method based on mood index, which is used in the process of human-computer interaction to carry out personalized modeling and description of the emotional state of a machine. Background technique [0002] In the interaction process between human and computer (or robot), the introduction of emotional factors can make the communication between them more natural and harmonious. Human-computer emotional interaction is to give computers the ability to observe, understand and generate various emotional characteristics similar to humans, and finally enable robots to interact naturally, vividly and emotionally like humans. [0003] For human-computer emotional interaction, in 1997, Professor Picard (Rosalind Picard) of the MIT Media Laboratory in the United States first proposed the concept of emotional computing. The research on emotional computing mainly includes emotional recognition, emotional re...

AI Technical Summary

Problems solved by technology

[0005] The above works have greatly promoted the research of emotion modeling, but most of these models cannot fully describe the dynamic process of human emotions, and there are many deficiencies in practice.

For example, cognition-based emotion models only consider the cognitive factors of emotion itself, and lack of consideration of non-cognitive factors such as personality; probability-based emotion models only deal with emotions from the perspective of pure probability, or by assuming a small amount of Kshirsagar’s multi-layered emotion model handles emotions in a vague way, which is not suitable for describing the complex emotions of human beings, nor can it link character and emotion well

Images