Indicator proposal for measuring regional political support for the electoral process on Twitter: The case of Spain's 2015 and 2016 general elections.

Research on electoral events in conjunction with social media provides opportunities to
describe an interesting phenomenon that can be analyzed using sentiment analysis techniques. The goal of
the study is to analyze the support of political parties during electoral periods from Twitter comments
including 250,000 tweets regarding the Spanish general elections of 2015 and 2016. Text mining and
natural language processing techniques enable information analysis, and the methodology emphasizes good
practices for large-scale data collection retrieved from Twitter through a quantitative analysis of text
collection written in the Spanish language. After information extraction obtained in three Spanish regions,
defined by geolocation, as well as feature selection based on keywords of the main four political parties, we
conducted an in-depth examination of Twitter users´ support during the course of the election. By
weighting the tendency of tweets, we were able to obtain a proposed indicator of support: the positiveness
ratio (PR). The results suggest that the PR is a feasible barometer to demonstrate measurable patterns of
support tendency regarding political parties and the users' behavioral activity to track their affinity on
Twitter. Findings indicate a consistent support behavior by users towards traditional parties and an
optimistic users’ behavior regarding the emerging political parties.