A Deep Learning based Self-Assessment Tool for Personality Traits and Interview Preparations

摘要

Many people face difficulty in analysing their own personality and to see whether they fit a particular job profile. Analysing our personality is very crucial, especially as a part of preparing for various types of interviews, as our responses reflects how we think and act, thus imprinting our first impression on the panelists. Studying personality traits has been proved to be an emerging stream using Machine Learning and Artificial Intelligence. Our idea is to create a platform to identify the personality traits of an individual and provide aid in suggesting changes, if required, in these traits. Our aim is to provide a helping hand for analysing how a person performs in various types of interviews like Video interview, Personal Interview, Group Discussions, etc. to ensure high-geared performance in final interviews. In our approach, we have used the Natural Language Processing (NLP) algorithm to analyse user's input in Group Discussion module, so as to provide additional context to the user. Sentiment Analysis of user's responses in Scenario Based Questions module results in how affirmative or negative the user's response is with respect to the expected solutions. For Video and Telephonic Interview modules, we have used MobileNet architecture and CNN algorithm to predict user's confidence level based on his/her facial expressions and voice modulation