Lunch at 12:30pm, talk at 1pm, in 148 Fitzpatrick

Title: Neural Models for Source Code Summarization

Abstract: Source code summarization is the task of writing short, natural language descriptions of source code. These descriptions form the backbone of a vast majority of software documentation used by programmers, and are usually encoded as code comments or other specially-formatted metadata. Even a brief summary such as ‘initializes microphone for web conference’ can tell a programmer a lot about what a section of source code does. But a conflict exists because even as programmers seek out high quality summaries in documentation, they are notorious for writing low quality summaries about their own code. This conflict has made code summarization a tempting target for automation. Strong research investment is now beginning to bear fruit, and the dream of automatic generation of code documentation is coming within our reach. In this talk, I will discuss neural models of code summarization, which form the current research frontier. I will discuss my own lab’s efforts and how they push this frontier.

Bio: I am an Associate Professor of Computer Science at the University of Notre Dame. I completed my Ph.D. at William & Mary, advised by Denys Poshyvanyk. My main research interests are at the intersection of software engineering and natural language processing, focused on software documentation generation.