International Journal of All Research Education and Scientific Methods (IJARESM), ISSN: 2455-6211 Volume 10, Issue 5, May-2022, Impact Factor: 7.429, Available online at: www.ijaresm.com IJARESM Publication, India >>>> www.ijaresm.com Page 1041 Personality Assessment Tool Using Ai for Friend’s Recommandation on Social Media Platform Sonali Dhokale 1 , Sushant Thorat 2 , Avinash Singh 3 , Shankar Tile 4 , Prof. Rahul Raut 5 1,2,3,4 U.G. Scholar, Department of Information Technology, Sandip Institute of Technology and Research Centre, Nashik, Maharashtra, India 5 Asst. Professor, Department of Information Technology, Sandip Institute of Technology and Research Centre, Nashik, Maharashtra, India --------------------------------------------------------------------*****************------------------------------------------------------------- ABSTRACT A Personality Assessment Tool Using AI For Friend’s Recommandation On Social Media Platform(Friend Suggestion System) for Social Networks uses the lifestyle of user to suggest friends. Many social networking sites recommend friends, items, books. For e.g. Facebook suggest the friends based on social relationship those who share common friends. Proposed system uses the lifestyle of user to recommend friend instead of social graph. The lifestyle of user can be determined from the user’s daily activities that he performs every day. Based on the similarity of lifestyle between the users the friend matching graph is drawn. The friend matching graph is generated in tabular form. System analyze friend matching graph to find out which users are more similar .Based on that users are recommended. System also allows user to give feedback about recommended friends.The emergence of social networking has led people to stay connected with friends, family, customers, colleagues or clients. Social networking can have social purposes, business purposes or both through sites such as Facebook, Instagram, LinkedIn, Twitter and many more. Recently, a large active social involvement have been seen from all eche- lons of society which keeps the friend circle increasing than never before.But, the friend suggestions based on one's friend list or profile may not be appropriate in some situations. Considering this problem, in this paper, a Friend Suggestion System, FA Finder (Friend Affinity Finder) based on 5 major dimensions (attributes): Agreeableness, Conscientiousness, Extraversion, Emotional range and Openness is proposed. This will help in understanding more about the commonalities that one shares with the other on the basis of their behaviour, choices, likes and dislikes etc. The suggested list of friends are extracted from the People Database (containing details of the 5 dimensions of different people) by deploying the concept of Hellinger-Bhattacharyya Distance (H-B Distance) as a measure of dissimilarity between two people.The Ocean Model is used to find friend behavior matching.For Behavior Classification Used Navie Bayes and Decision Tree.This study examines the relationship between use of social media, a popular online social network site, and the formation and maintenance of social capital. Keywords: AI(Artificial Intelligence, Naïve Bayes, Decision Tree, Big Five personality Traits. Social networking friend Agreeableness, Conscientiousness, Extraversion, Emotional range, Openness. INTRODUCTION A Friend Suggestion System for Social Networks uses the lifestyle of user to suggest friends. Many social networking sites recommend friends, items, books. For e.g. Facebook suggest the friends based on social relationship those who share common friends. Proposed system uses the lifestyle of user to recommend friend instead of social graph. The lifestyle of user can be determined from the user‟s daily activities that he performs every day. Based on the similarity of lifestyle between the users the friend matching graph is drawn. The friend matching graph is generated in tabular form. System analyze friend matching graph to find out which users are more similar .Based on that users are recommended. System also allows user to give feedback about recommended friends.Distanced from friends in olden days, people stood no chance in finding and getting connected to their friends again. Time passed and development of new ways of communication gave birth to internet. This constant pursuit of reliable connectivity gradually led to the development of social media. In recent times, social media plays a very crucial role in the _eld of communication. One of the important entities is finding or making new friends on social media. But, it is tougher than it seems because there are great chances of mismatch between the traits of different people in terms of behavioural attributes. With the growth of social media, social networking sites are facing the challenge of developing a proper and stable approach for recommending friends to the users. In order to overcome this problem and provide a more justified way of suggesting friends, a simple methodological web application has been developed to determine a good suggestive set of friends based on the attributes used: Agreeableness, Conscientiousness, Extraversion, Emotional range and Openness. In this paper, different users and their attribute ratings prime to define his/her behavioural qualities are considered.Social network