Reflecting the Tehran International Book Fair on X: A Sentiment Analysis of Users During the Event

Purpose: Driven by the Emotional Intelligence Theory, this study aims to identify and cluster the core discursive themes among X platform users during the 36th Tehran International Book Fair (2025). It specifically investigates the users' emotional and behavioral manifestations toward each topic in the virtual ecosystem. Crucially, the research endeavors to analyze the alignment between users' sentiment polarities and their subsequent digital interaction rates in response to this cultural phenomenon.
Method: The research dataset comprised 667 Persian tweets, manually collected during the timeframe of the 36th Tehran International Book Fair. Data preprocessing was conducted using the Parsivar library, an open-source natural language processing toolkit for Persian. Topic modeling was performed using the BERTopic model to identify semantically coherent clusters within the tweet content. Sentiment analysis was applied to evaluate users’ emotional orientations toward the identified topics. Furthermore, user engagement metrics, including views, likes, and retweets, were examined in relation to sentiments and topics.
Findings: After identifying and removing tweets classified as noise due to lack of content cohesion, the BERTopic model identified 11 main and distinct topics. The most frequent topic was related to the experience and intent of visiting and interacting at the fair, accounting for 17.8% of all tweets. Sentiment analysis revealed that the predominant emotional valence of tweets was positive; however, the topic of 'censorship and challenges related to content and supply restrictions' predominantly reflected negative sentiment. Tweets with negative and neutral emotional valence garnered higher engagement among users. Notably, the topic related to "Sacred Defense literature and books" received the highest engagement. Findings within the emotional analysis layer indicate that users with high emotional intelligence utilized the X platform to express empathy or offer constructive criticism. Conversely, challenging stimuli such as censorship activated users' emotional self-regulation mechanisms and triggered negative emotional contagion, leading to a dramatic surge in engagement rates.
Conclusion: Based on the research findings, valuable insights were provided for the event organizers and participating publishers. Understanding the discussed topics and their associated sentiments enables data-driven decision-making to enhance visitor satisfaction, address public concerns such as censorship and sales prohibitions, and optimize communication strategies for future editions of the Tehran International Book Fair. Moreover, the identified engagement patterns underscore the importance of adopting innovative approaches in media communication.These results highlight the necessity for cultural authorities to implement organizational emotional intelligence management to recalibrate public emotional responses during media crises.