Orestis Melkonian

Konstantinos Kogkalidis, Orestis Melkonian, Jean-Philippe Bernardy

Agda is a dependently-typed programming language and a proof assistant, pivotal in proof formalization and programming language theory. This paper extends the Agda ecosystem into machine learning territory, and, vice versa, makes Agda-related resources available to machine learning practitioners. We introduce and release a novel dataset of Agda program-proofs that is elaborate and extensive enough to support various machine learning applications -- the first of its kind. Leveraging the dataset's ultra-high resolution, which details proof states at the sub-type level, we propose a novel neural architecture targeted at faithfully representing dependently-typed programs on the basis of structural rather than nominal principles. We instantiate and evaluate our architecture in a premise selection setup, where it achieves promising initial results, surpassing strong baselines.

I. S. W. B. Prasetya, Craig Q. H. D. Leek, Orestis Melkonian et al.

There are many benefits in providing formal specifications for our software. However, teaching students to do this is not always easy as courses on formal methods are often experienced as dry by students. This paper presents a game called FormalZ that teachers can use to introduce some variation in their class. Students can have some fun in playing the game and, while doing so, also learn the basics of writing formal specifications in the form of pre- and post-conditions. Unlike existing software engineering themed education games such as Pex and Code Defenders, FormalZ takes the deep gamification approach where playing gets a more central role in order to generate more engagement. This short paper presents our work in progress: the first implementation of FormalZ along with the result of a preliminary users' evaluation. This implementation is functionally complete and tested, but the polishing of its user interface is still future work.