FLOC 2018: FEDERATED LOGIC CONFERENCE 2018
PaMpeR: A Proof Method Recommendation System for Isabelle/HOL

Authors: Yutaka Nagashima and Yilun He

Paper Information

Title:PaMpeR: A Proof Method Recommendation System for Isabelle/HOL
Authors:Yutaka Nagashima and Yilun He
Proceedings:Isabelle Papers
Editor: Makarius Wenzel
Keywords:Proof Method Recommendation, Machine Learning, Proof Mining, Regression Tree, Isabelle/HOL
Abstract:

ABSTRACT. Deciding which sub-tool to use for a given proof state requires expertise specific to each ITP. To mitigate this problem, we present PaMpeR, a proof method recommendation system for Isabelle/HOL. Given a proof state, PaMpeR recommends proof methods to discharge the proof goal and provides qualitative explanations as to why it suggests these methods. PaMpeR generates these recommendations based on existing hand-written proof corpora, thus transferring experienced users' expertise to new users. Our evaluation shows that PaMpeR correctly predicts experienced users' proof methods invocation especially when it comes to special purpose proof methods.

Pages:17
Talk:Jul 13 16:30 (Session 88C: Isabelle 4)
Paper: