Abstract:
"Numerous factors, such as a person's physiology, behaviour, and environment, have an impact on emotions. However, the majority of techniques for detecting mood have, up to now, depended on the interactions of a small number of scattered research components. This research examines a person's mood using physiological responses, behavioural tendencies, and a life-logging component. After preprocessing the data, multiple regression analysis was done to seek correlations between the variables. It has been shown that interactions between psychological arousal and valence levels strongly influence them. The findings of this study are as follows: The links between arousal and positive mood levels, as well as how these levels are impacted in real-world situations, were examined using data from people's life logs. With the help of the suggested wearable sensors, we are also able to build causal networks and identify the best coping mechanisms for each stressful circumstance that arises in daily life. This effort seeks to automate the mood-detection mechanism in order to address the issue. Previous work was thoroughly analysed to identify any unfulfilled gaps in the area of automated mood identification. Gaps in the domain of automated mood recognition are filled by the XGboost algorithm approach. Compared to earlier investigations, the findings of this one were more reliable. Future research using this strategy could include techniques for detecting mood changes that weren't included in this study. The results of this research were used as evidence of what had been effectively completed.
"
