Natural language processing of spatially crowdsourced data in petroleum revenue management

Abstract

AbstractIt has been over a decade of petroleum revenue utilisation in Ghana. Yet, there is a dearth of research on Ghanaians’ sentiments on petroleum revenue management in Ghana. However, research indicates that investigating citizens’ sentiments and addressing their grievances could prevent conflicts and promote better revenue utilisation in natural resource-rich countries. So, this study investigated Ghanaians’ sentiments about petroleum revenue management and its contribution to the Free Senior High School (SHS) programme in Ghana through an online survey. The study employed the quantitative approach in which the data was gathered through an online survey questionnaire and analysed using natural language processing techniques. The results show that the participants had negative sentiments about petroleum revenue management and the Free SHS programme in Ghana. However, they trust the managers and anticipate better revenue management in the future. The study recommends that the government should consult broadly with all stakeholders regarding petroleum revenue management to avoid potential conflicts. The article concludes that petroleum revenue managers can combine spatial crowdsourcing and natural language processing to extract citizens’ opinions at specific locations for better revenue management.