Understanding Colonial Legacy and Environmental Issues in Senegal Through Language Use

Abstract

Language is a reflection of issues and value systems of a society. This study tries to understand sensitive public issues in Senegal through language use. To this end, we utilize word embeddings, a numerical word representation, to analyze concepts, connotations, and nuances of several words. State-of-the-art machine learning methods can effectively extract the word embeddings from a collection of texts. Since people in different societies possess different mindsets and language uses, comparing semantic differences of words in different corpora is an efficient way to draw cross-cultural insights and implications. In this study, we extract word embeddings from Senegalese newspapers and French Wikipedia and then compare the results to identify different word sentiments in Senegalese cultures to understand the past, present, and future of the country.