Bloomberg has announced the creation of BloombergGPT, the first large language model built specifically for finance. Developed in collaboration with Johns Hopkins University, Bloomberg’s machine learning system was trained on a vast dataset combining general-purpose and financial-related text, giving it a unique ability to understand the complexities and nuances of the financial world. BloombergGPT outperformed similar models on financial task benchmarks, and its creators predict it will transform natural language queries from financial professionals into valid Bloomberg Query Language. The model will also improve financial NLP tasks, such as sentiment analysis, named entity recognition, news classification, and question answering, among others.
Mark Dredze, an associate professor of computer science at Johns Hopkins University, was a part of the team that developed Bloomberg’s AI system. Dredze is also the inaugural director of research (Foundations of AI) in the new AI-X Foundry at Johns Hopkins. He emphasized the importance of developing domain-specific machine learning models that can understand the unique terminology and complexities of their respective spheres. He added that such models will unlock new opportunities and transform existing tools such as Bloomberg Query Language.
Headquartered in New York City, Bloomberg is a financial data and media company that provides global business news in real-time, analytics, and electronic trading tools for both institutional and retail investors. With over 325,000 subscriptions, Bloomberg Terminal is the firm’s flagship product for financial professionals and is widely regarded as the industry-leading platform for data analysis and decision-making.
Mark Dredze, an associate professor of computer science at Johns Hopkins University’s Whiting School of Engineering, specializes in natural language processing and machine learning. His research focuses on developing technology that enables computers to understand and interact with human language. He has authored over 150 publications on machine learning and natural language processing and is one of the world’s leading experts in the field.
