Multi-Agent Aspect Level Sentiment Analysis in CRM Systems

Customer Relationship Management (CRM) became the best practice for any business that wishes to create, develop and enhance the customer value and implicitly the business shareholders value. Businesses became more aware that in the long term beyond the first sale customer retention is of crucial importance. However, in most cases, the first sale creates the first impression of the business. Being able to manage the customer expectations through aspect level sentiment analysis and proper guidance towards the first purchase, can make the difference between a strong retention rate and a weak retention rate. In this paper we present an approach for designing a multi-agent expert system using product aspect level sentiment analysis. The goal is to ease the conversion of a prospect to a customer by giving proper recommendations to accelerate the sale. Aspect level sentiment analysis takes into account not only the overall sentiment of the interaction but also the granular sentiment on the feature level of the products to be sold like for example price or quality. The multi-agent technology extends the CRM Systems and provides scalability, robustness and simplicity of design. Furthermore a prototype was developed and its design and results are presented and discussed.