Analysis of X users' opinions about political parties during the 2023 general elections period using text mining

This study aims to measure the level of public sentiment towards political parties in Turkey based on social media posts using sentiment analysis, one of the text mining methods. In this context, tweets shared through X were used. The study consists of two parts: theoretical and practical. In the theoretical part, literature reviews were conducted and the concepts of social media analysis, data mining, text mining and sentiment analysis were discussed. The research identified the five most debated political parties in the context of the 2023 general elections in Turkey. For these parties, more than seventy thousand tweets about five political parties on X were collected and 1250 of them, 250 for each party, were selected. The data was manually cleaned and labeled as positive, neutral and negative for sentiment analysis. As a result of the analysis, it was observed that both algorithms had similar accuracy rates in sentiment analysis, but in some cases, there were significant differences. As a result, the distribution of sentiment in tweets about political parties in Turkey varied by party and sentiment category. This paper is an important reference for future studies in the field of social media and politics and provides data that can guide researchers in the social media analysis of political parties.