22 March: DCLRS -- Magdalena Kacmajor, Friday, March 29 16:00 (TCD Arts
Dublin Computational Linguistics Research Seminar: Index of March 2019 | Dublin Computational Linguistics Research Seminar - Index of year: 2019 | Full index
Friday of next week (March 29, 2019), at 16:00, in room 3074 of the
Arts Building (TCD), Magdalena Kacmajor will speak.
Title:
Applying neural language model and neural translation model to source code
Abstract:
Software code is written not only for the machine, but also for
software developers who will be reviewing, extending or maintaining
the code. Thus writing software is an act of communication between
humans. The majority of software that is now available (in huge
amounts) from open source online repositories is simple and
repetitive, and research shows that statistical properties of source
code corpora can be predicted using statistical language models
developed for Natural Language Processing.
This talk concentrates on the use of neural language model for
predicting source code tokens, and the application of well-known
neural translation model to source code generation. After presenting
the general approach, we discuss main challenges in this area of
research, and suggest partial solutions to some of these
challenges. We also present the results of some of our experiments on
generating automated test code from pseudo-natural language
descriptions. The talk will draw on the work reported in (Kacmajor and
Kelleher, 2019).
Kacmajor, M. and Kelleher, J.D., 2019. Automatic Acquisition of
Annotated Training Corpora for Test-Code Generation. Information,
10(2):66-82. doi:10.3390/info10020066
ABOUT THE SPEAKER
Magdalena Kacmajor is an applied researcher within IBM Ireland
Innovation Exchange team, and a PhD student at TU Dublin. She is
involved in a Horizon 2020 project called ElasTest where she works on
developing cognitive engines for automated test generation and
recommendation.
------
The Dublin Computational Linguistics Research Seminar series, hosted
this year by the Trinity Centre for Computing and Language Studies, is
a cooperation among Trinity College Dublin, Dublin City University,
University College Dublin and the Dublin Institute of Technology, a
long standing collaboration which overlaps with the SFI CNGL/ADAPT
centres.
www.scss.tcd.ie/disciplines/intelligent_systems/clg/clg_web/DCLRS
Trinity Centre for Computing and Language Studies
www.scss.tcd.ie/CCLS
Dublin Computational Linguistics Research Seminar - Index of March 2019 | Index of year: 2019 | Full index