Modeling Person’s Creditworthiness over Their Demography and Personality Appearance in Social Media

Abstract

Financial institutions currently use credit history to determine whether to grant creditors credit. However, companies such as P2P Lending has a data shortage, especially credit history data, so innovative credit models emerge to improve the ability to assess creditors. Along with technology development, we have the opportunity to extract data from social media. This study uses social media data to create models for assessing creditworthiness. We collect data from social media and then process it using the credit scoring scorecard, linear correlation formula, credit scoring model weight composition, and threshold according to expert judgments. We find that by using a greater weight of the demographic attributes, we receive more data in the good credit category. This research on establishing model combinations contributes to assisting and making it easier for lenders to assess creditors using available data in a more practical way.