Quantifying fan engagement in sports using text analytics

• Published in: Journal of data, information and management (Online) Vol. 3; no. 3; pp. 197 - 208
• Main Author: H. Zadeh, Amir
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.09.2021; Springer Nature B.V
• Subjects: Algorithms; Artificial Intelligence; Automation; Business and Management; Business communications; Case studies; Classification; Computational Intelligence; Data collection; Data mining; Digital marketing; Digital media; Engineering; Expenditures; Information systems; Language; Mathematical analysis; Mathematical models; Original Article; Research methodology; Sentiment analysis; Social networks; Teams; Case study; Sentiment analysis; Social media; Opinion mining; University athletics; College sports; Fan engagement
• ISSN: 2524-6356; 2524-6364
• DOI: 10.1007/s42488-021-00052-4

The advent of digital communications has proliferated the engagements between customers and businesses of all varieties including university athletic programs. Interaction and engagement through social media content play a critical role in developing the relationship between fans and their favorite colligate teams. In this study, we reviewed the existing literature pertaining to the use of sentiment analysis and content categorization for fan engagement in the sports industry. Dozens of sources were examined, and their methodologies were explored. We present an analytic framework that can be used by sports organizations in their efforts to harness the power of AI and social media. The framework encompasses multiple stages related to textual data: data collection, data preparation, sentiment mining, and content categorization. In particular, this study demonstrates the use of text mining and sentiment analysis to provide athletic departments with more efficient and effective data understanding. In turn, this process will yield improved fan engagement to scale without increased expenditures. Using the textual data gathered from social media for a Basketball team at a major university in the United States, multiple analytical models were created using several different text mining packages, each one seeking to classify the polarity of the fan comments being examined. The study explored the possibility of classifying comments as positive or negative at the statement level. Statements were further categorized according to the subject matter of the comment. Inconsistencies were found between what the models identified and fan sentiment. Updating these models and the use of more effective text mining algorithms resulted in improved performance. Ultimately, it was determined that text mining and sentiment analysis models would be capable of performing the necessary analysis. Implications for research and practice are discussed.