Josh Dowdy obtained his master’s in Electrical and Computer Engineering from Mississippi State University in 2018. Throughout his master’s degree, he focused on machine learning, pattern recognition, and computational intelligence methodologies and algorithms related to image and signal processing. Josh was a member of the Research and Development team at Babel Street from January 2018 to December 2024 and has worked on various projects related to publicly available information processing. During his employment at Babel Street, he gained significant experience and domain knowledge in extracting valuable and relevant information from text using state-of-the-art natural language processing techniques. Throughout his employment at Babel Street, Josh has been part of projects involving document clustering, anomaly detection, and topic modeling and has led the creation of a geoinferencing system. He was also extensively involved in research utilizing large language models (LLMs) for risk/event detection. He is currently awaiting a decision on his application to the Ph.D. program of the Electrical and Computer Engineering department at Mississippi State University. As of January 2025, Josh has started working at Camgian as a Senior Data Scientist.
Contribution to Project: Josh has years of experience analyzing and extracting relevant information from large volumes of textual data composed of a variety of sources and languages. He has created, curated, and maintained pivotal databases and procedures to meet the requirements of a variety of tasks including topic modeling, document clustering, and anomaly detection. His expertise gained from working in an industry where customers rely on the results of an advertised product will be very useful as this project aims to provide knowledge and understanding of the COPE-ID dataset with hopes of being extended to support any dataset of its nature in the future. Josh has created all backend functionality and integrated Elasticsearch which allows a user of the platform to search for a variety of terms, topics, and sources across millions of documents within milliseconds. He has also worked closely with Chris and Jacob to establish a normalized data structure for the documents to ensure quality and accuracy.