Text Mining A Decade Of Focal Development Trends In An African Country

Abstract

This paper identified the emerging developmental trends of Ghana, an African country, based on online and offline textual, video and audio data sources from 2008 to 2018. In all, about 9,501 text files were collected and where needed, transcribed from these sources and content analyzed by series of text mining processes to bring out the convention for all the combined years. Topic modeling was the main adopted text mining technique. We initiated the topic modeling with the term frequency approach. The results from the mining processes were conceptually presented in relative frequency table and visually summarized as word clouds to map meaningful findings that were logically and humanly understandable. Through this approach, 45 major developmental themes emerged of which education seems to be the most frequently discussed trend in Ghanaian politics and government. In conclusion governments can detect developmental layouts from textual data.