An Algorithm-Centric Approach to Enhance Business Process Compliance Management

The diversity of enterprise domains is one of the significant factors that add to the challenges confronting organizations when managing compliance, and it, therefore, contributes to the cost of compliance. In particular, there is an indication that the inadequacy of a general approach to analyzing compliance documents impedes adequate compliance management practice. Taking an algorithmic-centric approach to tackling this challenge, this paper aims at providing an automated system that extracts, classifies and analyses compliance management documents in different domains. This paper tested and proposed machine learning techniques and systematically outlined algorithms for efficient extraction, classification and analysis of dataflow, control flow, and time constraints in compliance management domain for different stakeholders. The designed system was implemented using software tools. The classification was done in python Programming Language Environment using Machine Learning Algorithms while the Extraction and Analysis algorithms were implemented in Java programming language. Data were obtained from Central Bank of Nigeria (CBN), Nigeria Deposit Insurance Corporation (NDIC) and Nigeria Communication Commission (NCC). In the banking domain the authors chose Microfinance Banking system because it is a banking system that are commonly used by most government workers. Federal University of Agriculture Microfinance Bank was specifically chosen as a representative of all Microfinance Banks since they are all being guided by the same rules. The algorithms and techniques were implemented and validated through the two case studies and afterward evaluated for its adoption in practice.