Conference: Symposium on Computation and Pragmatics

The growing popularity of chat-based language models and other artificial intelligence (AI) systems makes social aspects of language a key priority for AI research. This is because when humans use language, they consider a rich set of information, including their relationship with their listeners and their listeners’ beliefs and goals. This project advances the development of socially-aware language technology by building bridges between experts in language science (including linguists) and technology (including computer scientists). This is an important advancement because it fosters a better scientific understanding of human language use and how to implement those principles in AI systems. Broader impacts include activities that bring computer scientists and linguists together for collaboration and mentoring activities to train new researchers in this area. The field of pragmatics explores interactive, context-sensitive, and discourse-level aspects of language. Although there has been progress in computational modeling of pragmatics and a growing body of applied computational work, for the most part language scientists and computer scientists have been working independently. This conference provides opportunities for computational linguists and natural language processing researchers to share ideas. Conference topics reflect many aspects of pragmatics including informativity, perspective-taking, image captioning, discourse coherence, mutual knowledge in dialogue, and emotion analysis. To maximize attendance, the conference is co-located with two established scientific gatherings. Virtual attendance, participation, and presentation are supported by an established video platform. Events consist of both invited and contributed presentations in addition to a pop-up mentoring event for trainees and other scholars that supports access to computational experts outside of the home institution. Conference presentations are published in open access proceedings. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.