The Sentiment Analysis, from the approach of the present project, focused on the monitoring of the interviews, as free and in-depth conversations, recorded on video, as well as the analysis of the anecdotal accounts collected a priori, to evaluate the language and voice inflections, and thus quantify the attitudes, opinions, gestures and emotions related to the implementation of the public policy aimed at promoting citizen participation in the sustainable and sustainable development of their communities..
The project consisted basically of the development of the following activities:
1.- Data collection.
2.- Data cleansing.
3.- Preprocessing (removal of stopwords and special characters in the text).
4.- Exploratory Data Analysis.
5.- Training a model for sentiment prediction using AI (Deep Learning).
6.- Visualization of sentiments, through a Dashboard.
Project Outcome
One of the most representative aspects for this analysis is found in the lower right box of the Dashboard, where the Artificial Intelligence facilitates the "Visual History" of the feelings throughout the complete narrative, in the form of "Valued Transformation of the Feeling" with respect to the "Total Narrative Time", in other words, how the participant flowed, when externalizing his perception regarding the application of university policies for participation in community development..