Direct Selling Business Lead Prediction by Social Media Data Mining

Abstract

Business leads are new potential customers and networkers from the direct selling business point of view, which are the marketing backbone of the direct selling industry. People who work in the direct selling business always want to enrich their contacts to promote their business. Today, with the huge increase of using internet technology by most people in the world, and with their activity information available on social media websites, it is possible to discover more suitable models for predicting potential people to contact. This thesis investigates some suitable data mining solutions for building a business lead prediction system framework over available social media data to suggest new potential customers and agents for supporting direct selling business. The information on Facebook friends’ list provides the networkers with business leads to help them to promote direct selling marketing. This research uses Facebook transactions as a case study for social media based lead prediction data mining because of its wide global usage. A set of data mining methods and algorithms are investigated and compared in determining the most suitable option based on feature analysis and selection of the social media data. Extensive experiments demonstrate and justify the proposed lead prediction system framework for supporting direct selling marketing promotion.