An Automated Approach to Contractual Compliance Analysis in Public Bidding Contracts

Abstract

Government procurement and contracting regulations mandate that public sector agreements include specific compulsory clauses that define key elements such as the contract’s subject matter, financial value, payment conditions, and delivery terms. However, these contracts are frequently published in unstructured formats, including scanned PDF documents or texts drafted without standardized templates, which hinders automatic information extraction. The processing of such documents therefore necessitates the use of Natural Language Processing (NLP) techniques and substantial computational resources. This study proposes an automated workflow for the identification of mandatory contractual clauses in public procurement documents, with a particular emphasis on extracting essential information to support compliance verification with the new legal framework. The proposed approach employs NLP techniques in combination with supervised machine learning models. The findings demonstrate that automating this process offers significant advantages, including enhanced transparency in public administration, improved detection of inconsistencies and potential irregularities, and greater efficiency in contract auditing and monitoring.