Sentiment analysis of Twitter messages regarding the KFC company inthe first quarter in Latin America 2022

Abstract

It was proposed to answer if the machines can really analyze
the sentiments of the tweets, then the messages in Spanish on
Twitter that spoke of KFC were analyzed. The tweets were
captured every day in the time period of the first quarter of
the year 2022 from the Latin American region, later they
were analyzed by month and for each company mentioned
in the tweets, these came to add 39,269 messages for KFC.

We focused on discovering what were the feelings related to each
message left, for this reason the polarity of the feeling between
positive and negative was sought, the first being related to wellbeing,
happiness, and love, while the second polarity, negative
was related to discomfort, sadness, and hatred. After obtaining
the polarity, it remained to discover what its degree was, the high,
medium and low indicators were used, thus having the degrees:
high positives, medium positives, low positives, high negatives,
medium negatives, and low negatives. The term neutral or neutral
was used for unpolarized messages, not meaning a feeling, that is,
neutral feelings do not exist, it is only the result of the absence of
sufficient data to classify it in some polarity. Everything mentioned
was done through artificial intelligence, but considering that it was
sought to answer if the feelings of the text messages can really be
analyzed, that is why two different heuristics were used, Machine
Learning and Deep Learning, with them it was possible identify
the polarity and degree of sentiment of Twitter messages regarding
the KFC company in the first quarter in Latin America 2022.