Regresar

Identifying the Political Tendency of Social Bots in Twitter Using Sentiment Analysis: A Use Case of the 2021 Ecuadorian General Elections

Abstract:

Sentiment analysis of social network data increasingly represents the real political scenario of many countries, which has turned bots into a powerful tool of influence, mainly due to their high efficiency. This work analyzes the messages on Twitter during the 2021 Ecuadorian presidential elections to determine sentiments and bots detection. We obtained a sample of 35,242 tweets corresponding to each candidate’s first and second rounds. Our methodology consists of four phases: first, we perform data collection using the Twitter API; secondly, we pre-process the data; in the third phase, we perform sentiment analysis of the content of the tweets to understand their posture towards a candidate, and finally, we classify the users as bots or not. As a result, we discovered that bots and non-bots people on both sides had more positive feelings towards their respective candidates than unfavorable feelings against the other candidates.