Abstract: 

Machine learning has been widely incorporated into formal reasoning; however, its utilization in term rewriting—a field intrinsically linked to formal reasoning—remains relatively underexplored. Term rewriting serves as a mathematical abstraction of programs and provides a framework for analyzing various program properties, such as termination and complexity. This presentation offers a brief introduction to term rewriting and examines recent progress in integrating machine learning into this domain. In particular, it addresses automated strategy generation, neural-based strategy selection, and neural algorithms for property analysis in term rewriting.