This National Science Foundation (NSF) Computer and Information Science and Engineering (CISE) Project Grant (CFDA 47.070) award of $400,000 to Texas A&M University-Corpus Christi aims to develop an advanced hybrid machine learning pipeline to enhance the analysis of speech and detection of threatening speech online. The project focuses on three key aspects: 1) detecting the characteristics of individual speech posts using large language models, 2) understanding and mitigating the spread of threatening speech through graph neural networks, and 3) addressing the lack of comprehensive multilingual datasets, particularly for English and Spanish-speaking communities.
The project includes a $285,874 sub-award to Mississippi State University to support tasks on source localization for hate speech spreaders, interaction decomposition at the group level, and influence blocking strategies within social networks. The overall award seeks to advance speech detection research and create new multilingual datasets to support effective mitigation of online threatening speech.
Generated 3/4/25, 6:08 AM