AI Automation of B2B Event Management Using Predictive Workflows and Entity Matching

Abstract

There are many times consuming and complicated tasks that have to be done in business-to-Business (B2B) event management like attendee targeting, lead distribution and scheduling of the vendors. The given paper introduces the idea of an AI-based automation platform that can ease these activities with the help of predictive modelling and matching of entities. It involves a predictive workflow that is based on random forest to predict the interest and engagement of the attendees based on features name, email, company, role, past_ events_ attended, opened _ email, clicked link, registered and attended. Simultaneously, an entity matching component that applies NLP and fuzzy logic automatically links and eliminates duplicates between database contacts, which removes the necessity of manual cleansing. Data preprocessing consists of Label Encoding the categorical variable and normalizing the numerical through Standard Scaler. Real-world B2B event data evaluation indicated extremely high accuracies of 92%, precision of 0.89, recall of 0.90 and F1-score of 0.91. The automation is a big boost to better planning, matchmaking, predication, and lead conversion. The next iteration will have a real time feedback loop and multi-language entity recognition so as to be able to scale the global B2B events.