Machine Learning for Personality Analysis Based on Big Five Model

The proposed research attempts to emulate a statistical report making system, which takes into considerations, the activities of user and their behavior online by means of their interactions on varied array of social media platforms. It is possible that youngsters may come across incidents on Internet, which probably may be inappropriate for their age group or may push them towards certain erratic psychological behaviors. This study caters to such arising needs, for various individuals-young or old, alike, so as to keep a tab upon their own online activities through browsing history which may be directly/indirectly blend into their human characteristics. On social media, people express their likes, dislikes, thoughts, opinions, and feelings which sum up to be their own personality. This data (thoughts and opinions on social platform and browsing history) can be exponentially aggregated to identify user's personality traits. It can then be used for self-monitoring, parental monitoring, or for businesses who wish to hire employees based on their personality criteria, if approved by concerned users. For this study, we have used supervised machine learning algorithms like Naive Bayes and Support Vector Machines. We have evaluated their performance through the combinations of different feature extraction process like BOW, TF, and TF-IDF with each classifier. In conclusion, we have found that TF-IDF with SVM has the best performance.