Planned intervention: On Wednesday April 3rd 05:30 UTC Zenodo will be unavailable for up to 2-10 minutes to perform a storage cluster upgrade.
Published March 15, 2019 | Version v1
Preprint Open

Evaluation of Move Method refactorings recommendation algorithms: are we doing it right?

  • 1. Saint Petersburg State University
  • 2. JetBrains Research, Higher School of Economics
  • 3. JetBrains Research, Saint Petersburg State University

Description

Previous studies introduced various techniques for detecting Move Method refactoring opportunities. However, different authors have different evaluations, which leads to the fact that results reported by different papers do not correlate with each other and it is almost impossible to understand which algorithm works better in practice. In this paper, we provide an overview of existing evaluation approaches for Move Method refactoring recommendation algorithms, as well as discuss their advantages and disadvantages. We propose a tool that can be used for generating large synthetic datasets suitable for both algorithms evaluation and building complex machine learning models for Move Method refactoring recommendation.

Files

IwoR_19.pdf

Files (96.3 kB)

Name Size Download all
md5:fa347bc4e60617498016fb2a0e9d5b11
96.3 kB Preview Download