Improving relationship management in universities with sentiment analysis and topic modeling of social media channels

Online Social networking (OSN) platforms such as Facebook, daily have a massive number of users and content being created. Users of such services have the power to share opinions and influence others. This creates an interesting scenario, where brands and institutions can have a digital presence to interact directly with their target audience. Universities around the world are also using those platforms for reaching out their students and staff, acting as a new channel for services and public relations. Social Customer Relationship Management (SCRM) principles can be applied in this scenario, for universities, and help them on the management of this relationship with their public. Thus, this work aims to apply SCRM methodology for a university Facebook fan page, through techniques of Sentiment Analysis (SA) and topic modeling (TM) using Latent Dirichlet Allocation, in order to find topics people are complimenting or complaining about in their comments. The main goal is to improve SCRM processes through the results and insights provided by those techniques. The final discussion with the communications professionals from UFPA university has revealed the experiments and insights provided in this work are valuable for their social media (SM) management workflow.