Paper Title
Extracting Personality Features From User Generated Data For Recruitment Via Text Mining

Abstract
Social media has been used for different purposes under the name of big data to extract useful information. In the field of human resources social media is used manually by managers and operational staff especially for recruitment. On the other hand personality tests are some other instruments to evaluate candidates for hiring process. In this study, we used social media of 30 volunteers to find out their personality traits like dominant, compliance, steadiness, influence. We have interviewed and given them the D.I.S.C test and we have evaluated them with the answers they have given. After this manual assessment process with the permit of those volunteers, we have extracted their tweets from social media i.e Twitter. These text data have been converted into a data warehouse that can be processed with text mining tools. After that, clustering has been applied to the data. Clustering results have been cross validated with the manual assessment results. For each cluster, common key words are discovered to represent characteristics like dominant, compliance, steadiness, influence. Our study shows that, when social media entries or tweets are examined with text mining tools, some words may give a clue about the character of the user. Although this study has used texts produced on social media, any user generated data may be applicable for the test or analysis. Keywords- Social media; human resources; personality tests; big data; text mining.