Customer Profiling from Social Media Engagement using Latent Dirichlet Allocation (LDA) and sentiment Analysis Approach

Not scheduled
15m
Abstract for Research Paper Digital Platforms

Description

Social media is now an essential component of the daily life of consumers. People usually share their interest, thoughts on brands and companies through discussions, tweets and status. At present, companies are competing to attract and meet customer needs. For companies, managing customer relationship through social media engagement has become a significant part of digital marketing strategies. The modern customer has different needs, expectations and behaviours which ought to be managed differently by companies., Customer engagement on social networks helps to create relationship with customers, and also acts as quick and cost-effective marketing tool. Social Customer Relationships Management (SCRM) provides a two-way communication channel between customers and businesses through social media sites. SCRM is based on a model of customer engagement which requires strong partnerships and interactions. The purpose of this research study was to understand customer interactions with business using topic modelling. The study analysed customer engagement on Twitter of Four selected banks in Kenya. We apply unsupervised topic modelling of LDA and sentiment analysis to create a profile of different customers of selected banks in Kenya. We focus on interactions from a consumer-centric perspective, not focusing on specific firm channels. We conclude that the extracted latent models not only provide insight to the consumer behaviour but also can also improve any company’s Social Customer relationship management(sCRM) focused on different customer profiles
Keywords: Social Media; Customers, Sentiment analysis; Topic modelling, LDA, CRM, sCRM

Primary authors

Patrick Kanyi (United States International University Africa) Japheth Mursi (UKZN)

Presentation Materials

There are no materials yet.