Education is the most essential part of life. It nurtures so many skills in life. There are many obstacles that may occur during the course of education such as physical or mental health problems and learning disabilities. The affective states representing the human behavior are to be observed for the learners. The affective states can be represented in various dimensions including valence, arousal etc. There is always a need to sense the mood of students using those dimensions such as valence and arousal in social media to analyze the various factors that are acting as obstacles. On identifying the elements acting as obstacles it will be easy to mentor the students to prevent them falling into depression. The proposed mentoring model can be used to improve the student’s performance through mentoring by sensing their mood using valence and arousal factors. © 2017, Institute of Advanced Scientific Research, Inc. All rights reserved.