Sk-P: A neural program corrector for MOOCs

Yewen Pu, Karthik Narasimhan, Armando Solar-Lezama, Regina Barzilay

Research output: Chapter in Book/Report/Conference proceedingConference contribution

29 Scopus citations

Abstract

We present a novel technique for automatic program correction in MOOCs, capable of fixing both syntactic and semantic errors without manual, problem specific correction strategies. Given an incorrect student program, it generates candidate programs from a distribution of likely corrections, and checks each candidate for correctness against a test suite. The key observation is that in MOOCs many programs share similar code fragments, and the seq2seq neural network model, used in the natural-language processing task of machine translation, can be modified and trained to recover these fragments. Experiment shows our scheme can correct 29% of all incorrect submissions and out-performs state of the art approach which requires manual, problem specific correction strategies.

Original languageEnglish (US)
Title of host publicationSPLASH Companion 2016 - Companion Proceedings of the 2016 ACM SIGPLAN International Conference on Systems, Programming, Languages and Applications
Subtitle of host publicationSoftware for Humanity
EditorsEelco Visser
PublisherAssociation for Computing Machinery, Inc
Pages39-40
Number of pages2
ISBN (Electronic)9781450344371
DOIs
StatePublished - Oct 20 2016
Externally publishedYes
Event2016 ACM SIGPLAN International Conference on Systems, Programming, Languages and Applications: Software for Humanity, SPLASH Companion 2016 - Amsterdam, Netherlands
Duration: Oct 30 2016Nov 4 2016

Publication series

NameSPLASH Companion 2016 - Companion Proceedings of the 2016 ACM SIGPLAN International Conference on Systems, Programming, Languages and Applications: Software for Humanity

Other

Other2016 ACM SIGPLAN International Conference on Systems, Programming, Languages and Applications: Software for Humanity, SPLASH Companion 2016
CountryNetherlands
CityAmsterdam
Period10/30/1611/4/16

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Computer Science Applications
  • Software

Keywords

  • Code repair and completion
  • Language model
  • MOOCs

Fingerprint Dive into the research topics of 'Sk-P: A neural program corrector for MOOCs'. Together they form a unique fingerprint.

  • Cite this

    Pu, Y., Narasimhan, K., Solar-Lezama, A., & Barzilay, R. (2016). Sk-P: A neural program corrector for MOOCs. In E. Visser (Ed.), SPLASH Companion 2016 - Companion Proceedings of the 2016 ACM SIGPLAN International Conference on Systems, Programming, Languages and Applications: Software for Humanity (pp. 39-40). (SPLASH Companion 2016 - Companion Proceedings of the 2016 ACM SIGPLAN International Conference on Systems, Programming, Languages and Applications: Software for Humanity). Association for Computing Machinery, Inc. https://doi.org/10.1145/2984043.2989222