“Mood Detector”

Abstract

Understanding human thoughts since times always remained a mysterious challenge for the scientific discipline as is the circumstance with the emotions of human beings. The numerous techniques of Emotion detection has already been discovered, one among which we here have discovered is the detection of emotion which is here done using IoT and Machine learning technique. This paper is being proposed to present the design and implementation of a Mood Detector application, which has been designed to detect the mood and emotional state of a person by examining the triad physical constraints (temperature, pulsate, motion and skin electro-conductance) by using a machine learning algorithm which is trained with data provided by the mood detector application developer. This application has been tested redundantly unless and until the results generated by the learning algorithm have been validated to 100%, thus affirming that the machine learning algorithms provides the accurate results. This application also coordinates a music recommender framework, which recommends the user to listen to the vague playlists, which has been designed to the recognized mood. In this paper, we design a probabilistic data collection mechanism and on the collected data we perform a correspondence analysis. Finally we design a statistical model to anticipate the human temperament and recommend a music playlist in accordance with their current temperament.