Automated Error Correction for Concurrent Logic Programs

Title: Automated Error Correction for Concurrent Logic Programs

Abstract: This research presents Kima, an automated error correction system for concurrent logic programs. Kima uses a constraint-based analysis technique called mode/type analysis to detect and correct errors in program code. The key contribution of this research is an algorithm that efficiently finds the minimal inconsistent subset of mode/type constraints, which can pinpoint the possible sources of errors in the program. Kima uses this information to propose a small number of alternative programs that include the intended program, making the error correction process more efficient. The research demonstrates the effectiveness of Kima through quantitative experiments.

Main Research Question: How can we develop an automated error correction system for concurrent logic programs that is efficient and effective in detecting and correcting program errors?

Methodology: The research uses a constraint-based analysis technique called mode/type analysis, which is applied to concurrent logic programs written in a language called Moded Flat GHC. This technique allows for the reconstruction of types and modes from program text, which helps in detecting inconsistencies in the program. The research proposes an algorithm that finds the minimal inconsistent subset of mode/type constraints, which can pinpoint the possible sources of errors in the program.

Results: The research implemented Kima, an automated error correction system for concurrent logic programs, which uses the proposed algorithm to efficiently find and correct program errors. The experiments showed that Kima was able to correct a large number of errors in the program, making it a practical and effective tool for debugging concurrent logic programs.

Implications: The research has important implications for the field of automated debugging. It demonstrates that constraint-based analysis techniques can be used to develop efficient and effective error correction systems for concurrent logic programs. The research also provides a practical solution for debugging programs that are not completely constructed, which is a common scenario in software development.

Link to Article: https://arxiv.org/abs/0012007v3 Authors: arXiv ID: 0012007v3