We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: MISIM: A Neural Code Semantics Similarity System Using the Context-Aware Semantics Structure

Abstract: Code semantics similarity can be used for many tasks such as code recommendation, automated software defect correction, and clone detection. Yet, the accuracy of such systems has not yet reached a level of general purpose reliability. To help address this, we present Machine Inferred Code Similarity (MISIM), a neural code semantics similarity system consisting of two core components: (i)MISIM uses a novel context-aware semantics structure, which was purpose-built to lift semantics from code syntax; (ii)MISIM uses an extensible neural code similarity scoring algorithm, which can be used for various neural network architectures with learned parameters. We compare MISIM to four state-of-the-art systems, including two additional hand-customized models, over 328K programs consisting of over 18 million lines of code. Our experiments show that MISIM has 8.08% better accuracy (using MAP@R) compared to the next best performing system.
Comments: arXiv admin note: text overlap with arXiv:2003.11118
Subjects: Machine Learning (cs.LG); Software Engineering (cs.SE); Machine Learning (stat.ML)
Cite as: arXiv:2006.05265 [cs.LG]
  (or arXiv:2006.05265v6 [cs.LG] for this version)

Submission history

From: Shengtian Zhou [view email]
[v1] Fri, 5 Jun 2020 20:55:23 GMT (9161kb,D)
[v2] Mon, 15 Jun 2020 19:56:43 GMT (9161kb,D)
[v3] Tue, 6 Oct 2020 19:39:45 GMT (9509kb,D)
[v4] Fri, 9 Oct 2020 00:25:35 GMT (9509kb,D)
[v5] Mon, 22 Feb 2021 23:46:57 GMT (11982kb,D)
[v6] Wed, 2 Jun 2021 20:34:31 GMT (12258kb,D)

Link back to: arXiv, form interface, contact.