LAPS: Computing Loan Default Risk from User Activity, Profile, and Recommendations

The credit score is one variable in receiving a loan application from a bank or financial institution that provides credit/loan. Many factors determine whether a borrower gets the loan. One of them is through more valuable collateral than the loan that was proposed. However, this is not possible for borrowers to provide it. Personal data, job information, salary amounts, assets owned, and valuable documents are usually required to determine a credit score. We build a personal lending platform model based on the trustworthiness score called LAPS (Loan Risk score, Activity score, Profile score, and Social Recommendation score) borrower trustworthiness score. The borrowers' trustworthiness is an absolute requirement to ensure they can repay the loans and installments on time. We present the practical ways to select the best features from the Bank Marketing dataset. The feature selection of the dataset applies to blockchain applications. The advantage of LAPS is introducing recommenders' as guarantors to convince the lenders'/investors' and minimizes collateral by implementing a LAPS.