Pointing to Subwords for Generating Function Names in Source Code

11/09/2020
by   Shogo Fujita, et al.
0

We tackle the task of automatically generating a function name from source code. Existing generators face difficulties in generating low-frequency or out-of-vocabulary subwords. In this paper, we propose two strategies for copying low-frequency or out-of-vocabulary subwords in inputs. Our best performing model showed an improvement over the conventional method in terms of our modified F1 and accuracy on the Java-small and Java-large datasets.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
08/10/2021

Megadiff: A Dataset of 600k Java Source Code Changes Categorized by Diff Size

This paper presents Megadiff, a dataset of source code diffs. It focuses...
research
03/17/2020

Big Code != Big Vocabulary: Open-Vocabulary Models for Source Code

Statistical language modeling techniques have successfully been applied ...
research
02/06/2022

Featherweight Assisted Vulnerability Discovery

Predicting vulnerable source code helps to focus attention on those part...
research
05/21/2020

Java Decompiler Diversity and its Application to Meta-decompilation

During compilation from Java source code to bytecode, some information i...
research
04/06/2020

Embedding Java Classes with code2vec: Improvements from Variable Obfuscation

Automatic source code analysis in key areas of software engineering, suc...
research
08/29/2018

Mapping Language to Code in Programmatic Context

Source code is rarely written in isolation. It depends significantly on ...
research
08/19/2019

The Strengths and Behavioral Quirks of Java Bytecode Decompilers

During compilation from Java source code to bytecode, some information i...

Please sign up or login with your details

Forgot password? Click here to reset