Sentiment Analysis on the young people's perception about the mobile Internet costs in Senegal
Derguene Mbaye, Madoune Robert Seye, Moussa Diallo, Mamadou Lamine Ndiaye, Djiby Sow, Dimitri Samuel Adjanohoun, Tatiana Mbengue, Cheikh Samba Wade, De Roulet Pablo, Jean-Claude Baraka Munyaka, Jerome Chenal
TL;DR
The study investigates how young Senegalese perceive mobile Internet costs by collecting a large multilingual social-media corpus from Facebook and Twitter across four operators and applying a sentiment analysis pipeline that leverages translation to French plus a multilingual model to handle Wolof. It reveals general dissatisfaction with the price/quality balance, with operator-specific patterns (e.g., strong negative sentiment toward Orange’s pricing and relatively more positives for Promobile). The approach combines data collection, language identification for Wolof, word-cloud exploration, and cross-language sentiment classification to address low-resource language challenges. The work highlights regulatory and market implications and outlines future steps to create open Wolof sentiment resources to support more robust NLP in telecom contexts.
Abstract
Internet penetration rates in Africa are rising steadily, and mobile Internet is getting an even bigger boost with the availability of smartphones. Young people are increasingly using the Internet, especially social networks, and Senegal is no exception to this revolution. Social networks have become the main means of expression for young people. Despite this evolution in Internet access, there are few operators on the market, which limits the alternatives available in terms of value for money. In this paper, we will look at how young people feel about the price of mobile Internet in Senegal, in relation to the perceived quality of the service, through their comments on social networks. We scanned a set of Twitter and Facebook comments related to the subject and applied a sentiment analysis model to gather their general feelings.