Sentiment analysis for smart city projects in Africa. Codebase for our paper, Highlighting smart city mirages in public perceptions: A Twitter sentiment analysis of four African smart city projects.

V1 - Using project title as the search keyword.
V2 - Adjusted search incorparoting specific terms like 'smart city', 'smart cities' and the country the project is found in.
V3 - Same as V2 but without tweets from the official smart city project account. This is to eliminate bias. V4 - A redo of V3 with less preprocessing. VADER sentiment analysis works with patterns optimized for social media... too much pre=processing makes a tweet look less like a regular tweet and can hurt performance.