,
Associate Professor, Department of Information Technology, St. Joseph Institute of Technology, Chennai, Tamil Nadu, India
Rapid development of digital platforms has led to an abundance of data collected from customers, necessitating the use of intelligent methods that can be used to analyze behavior patterns and aid in making effective Customer Relationship Management (CRM) decisions. Current CRM artificial intelligence systems mainly aim at high prediction accuracy and rarely provide features like interpretable analysis, contextual reasoning, and interpretation. To overcome this challenge, the current research introduces the NeuroCRM-KM (Neuro-Symbolic Knowledge Mining) approach that can be used to analyze customers’ behavior patterns and aid in making decisions through neural learning, knowledge graph representation, and symbolic reasoning. In particular, the NeuroCRM-KM method involves collecting heterogeneous customer information such as transactions, demographics, customer reviews, and interactions, and then proceeding with data preprocessing, behavioral analysis using statistics, feature learning using neural networks, knowledge graph creation, and symbolic rule-based reasoning. The experimental evaluation performed on 5,000 customers' data shows that the suggested NeuroCRM-KM framework provides better results than those obtained with the help of machine learning and deep learning frameworks by achieving 96.4% accuracy, 95.8% F1 score, and 0.97 AUC. In addition, the ablation study has proven the significance of symbolic reasoning, knowledge graph integration, and neural learning elements for predicting performance and making it more interpretable. Finally, the result of the research is represented by the intelligent and explainable CRM decision-making system that allows for personal marketing, customer retention, and business strategy planning.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
0
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.