Sentiment analysis and mood detection on an android platform using machine learning integrated with internet of things

Abstract

Mental health is considered as one of the most sensitive topics of research and it is highly affected by an individual’s mood and sentiments. Social media has been proven to be one of the major catalysts in deterioration and fickleness of one’s mind. In this paper, we present an android application called “moody buddy” ingratiated with a heartbeat analyzing hardware kit which would detect and analyze the moods and emotions of an individual very close to accuracy. Mood recognition and sentiment analysis is a vast and complex area of research. Moreover, monitoring human emotions is found out to be one of the technically challenging aspects. So, in order to achieve the quality output of our research and testing work, we have taken help from artificial intelligence and Internet of Things domain. Here, we have considered the activity of the user on his/her social networking as a starting point of our research work. The concept of logistic regression is used in our software. In order to solidify our idea more, we are adding a hardware component which would monitor the heartbeat of the person and its modulation. In case of any abnormality examined in the heart rate, the questionnaire appears again. At the end, a cumulative of the hardware component’s results and software component’s would help us analyze and detect the current mood of the individual to very close to high accuracy value.