Paper published in a book (Scientific congresses, symposiums and conference proceedings)
Assessing the Generalizability of code2vec Token Embeddings
Kang, Hong Jin; Bissyande, Tegawendé François D Assise; David, Lo
2019In Proceedings of the 34th IEEE/ACM International Conference on Automated Software Engineering
Peer reviewed
 

Files


Full Text
ase19-code2vec.pdf
Author preprint (250.45 kB)
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
Code Embeddings; Distributed Representations; Big Code
Abstract :
[en] Many Natural Language Processing (NLP) tasks, such as sentiment analysis or syntactic parsing, have benefited from the development of word embedding models. In particular, regardless of the training algorithms, the learned embeddings have often been shown to be generalizable to different NLP tasks. In contrast, despite recent momentum on word embeddings for source code, the literature lacks evidence of their generalizability beyond the example task they have been trained for. In this experience paper, we identify 3 potential downstream tasks, namely code comments generation, code authorship identification, and code clones detection, that source code token embedding models can be applied to. We empirically assess a recently proposed code token embedding model, namely code2vec’s token embeddings. Code2vec was trained on the task of predicting method names, and while there is potential for using the vectors it learns on other tasks, it has not been explored in literature. Therefore, we fill this gap by focusing on its generalizability for the tasks we have identified. Eventually, we show that source code token embeddings cannot be readily leveraged for the downstream tasks. Our experiments even show that our attempts to use them do not result in any improvements over less sophisticated methods. We call for more research into effective and general use of code embeddings.
Disciplines :
Computer science
Author, co-author :
Kang, Hong Jin;  Singapore Management University > SIS
Bissyande, Tegawendé François D Assise  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT)
David, Lo
External co-authors :
yes
Language :
English
Title :
Assessing the Generalizability of code2vec Token Embeddings
Publication date :
November 2019
Event name :
34th IEEE/ACM International Conference on Automated Software Engineering
Event place :
San Diego, California, United States
Event date :
from 10/11/2019 to 15/11/2019
Audience :
International
Main work title :
Proceedings of the 34th IEEE/ACM International Conference on Automated Software Engineering
Pages :
1-12
Peer reviewed :
Peer reviewed
Focus Area :
Security, Reliability and Trust
Available on ORBilu :
since 23 January 2020

Statistics


Number of views
95 (8 by Unilu)
Number of downloads
142 (6 by Unilu)

Scopus citations®
 
49
Scopus citations®
without self-citations
44

Bibliography


Similar publications



Contact ORBilu