Proposing NLP algorithms for shariah screening platform (SHARSP)

Abstract

This research proposes an algorithm for Shariah Screening Platform to assist Shariah decision making process in the Islamic Financial Industry. It is designed based on mixed methodology approach which involves both quantitative as well as qualitative method for data collection and analysis. Qualitative approach is first done by collecting sample of bank product proposal and guidelines from the Central Bank of Malaysia for similarity algorithm development. The first test of similarity is done using Levenshtein Distance. Given the algorithm written in Python, this research tests two different sets of text data where the first one is sample of Tawarruq product proposal (query) with Tawarruq policy document (corpus) and the second set is the same Tawarruq proposal (query) with Shariah Governance Policy Document as corpus. A suitable algorithm is built for document similarity analysis using Python programming language. Two measures of similarity is adopted which are Levenshtein Distance and FuzzyWuzzy Package to compare two strings document. These algorithms then are used to develop a Shariah Screening Platform named as SHARSP. The second stage is quantitative approach where data is collected through online structured survey distributed among industry practitioners to observe their view on the proposed Shariah Screening Platform. Based on Technology Acceptance Model (TAM) and Technology Readiness Index (TRI) as a measurement tool and multiple regression analysis, majority of the respondents were in view that the proposed SHARSP is useful, and they were prone towards the industry innovation. Overall, outputs of this research are significant to Islamic financial industry in proposing new technology for document screening. It would also add to the existing literatures on Islamic finance especially in the adoption of NLP technology.

Keywords: Shariah Screening Platform, Islamic Financial Industry, TAM, TRI, NLP