Diya Saha

RegNLI: Detecting Online Product Misbranding through Regulatory and Linguistic Alignment

This work introduces RegNLI, a novel framework that formulates misbranding detection as a inference task between product claims and regulatory provisions, and builds a foundation for compliance-aware NLP systems and opens new directions for integrating formal reasoning with neural architectures in regulatory domains.