Language and Multimodal Models in Sports: A Survey of Datasets and Applications

Haotian Xia,Zhengbang Yang,Yun Zhao,Yuqing Wang,Jingxi Li,Rhys Tracy,Zhuangdi Zhu,Yuan-fang Wang,Hanjie Chen,Weining Shen
2024-06-18
Abstract:Recent integration of Natural Language Processing (NLP) and multimodal models has advanced the field of sports analytics. This survey presents a comprehensive review of the datasets and applications driving these innovations post-2020. We overviewed and categorized datasets into three primary types: language-based, multimodal, and convertible datasets. Language-based and multimodal datasets are for tasks involving text or multimodality (e.g., text, video, audio), respectively. Convertible datasets, initially single-modal (video), can be enriched with additional annotations, such as explanations of actions and video descriptions, to become multimodal, offering future potential for richer and more diverse applications. Our study highlights the contributions of these datasets to various applications, from improving fan experiences to supporting tactical analysis and medical diagnostics. We also discuss the challenges and future directions in dataset development, emphasizing the need for diverse, high-quality data to support real-time processing and personalized user experiences. This survey provides a foundational resource for researchers and practitioners aiming to leverage NLP and multimodal models in sports, offering insights into current trends and future opportunities in the field.
Computation and Language
What problem does this paper attempt to address?
The paper aims to systematically explore the integrated application of Natural Language Processing (NLP) and multimodal models in the sports domain, providing a comprehensive dataset and application review. Specifically, the paper addresses the following key issues: 1. **Dataset Classification**: The paper categorizes datasets into three types: language-based datasets, multimodal datasets, and convertible datasets. These datasets are used for text or multimodal tasks (such as text, video, audio), and single-modal datasets (such as video) can be converted into multimodal datasets by adding annotations. 2. **Application Areas**: The paper details various applications of NLP and multimodal models in the sports domain, including match prediction and analysis, hate speech detection, named entity recognition, news summarization, fan interaction, sports understanding, medical applications, and educational applications. 3. **Future Directions and Challenges**: The paper discusses the future development directions and challenges of NLP and multimodal models in the sports domain, emphasizing the need for more high-quality and diverse data to support real-time processing and personalized user experiences. Through this review, the paper provides a foundational resource for researchers and practitioners, helping them better utilize NLP and multimodal models in sports research and practice.