Research on Contract Audit Automation System Based on Neural Networks

An audit of a contract is a type of review that takes place intending to ascertain whether one or both parties of a contract have complied with the provisions of a contract or not. New contract risks and volumes in various businesses have led to increased audit requirements in terms of ability, efficiency, and sophistication. Standard contract auditing can be described as tedious, manual, imprecise, and time-consuming. This research work also introduces the Contract Audit Automation System, which has incorporated the use of a neural network. In this study, we introduce a new beetle swarm-optimized adaptive long and short-term memory model (BSO-ALSTM), which can improve the efficiency of contract auditing. Data about contracts are collected from different companies including in the fields of finance and purchasing. The proposed methodology involves further analysis of the contract text and extraction of relevant phrases through natural language processing (NLP). Tokenization was used in analyzing the data to help isolate important aspects of the data. The proposed method is compared with other traditional algorithms to analyze its performance. The results attest that the proposed method has offered enhancements in audit speed and accuracy as compared to other algorithms. This research focuses on ways through which the technologies can be adapted to support contract management and auditing, a solution that can be replicated and adapted to meet different industry requirements.