Share Email Print
cover

Proceedings Paper

Paraphrase generation and evaluation: a view from the trenches
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

In this paper we evaluate the current state of the art in natural language paraphrase generation using deep learning methods. The focus is put on the entire modeling pipeline from data gathering up to model evaluation. Specifically, we list the publicly available datasets suitable for this task, assess their quality and discuss procedures connected with data preparation and model training. Finally, we discuss problems related to the currently used evaluation approaches.

Paper Details

Date Published: 6 November 2019
PDF: 10 pages
Proc. SPIE 11176, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2019, 111761M (6 November 2019); doi: 10.1117/12.2535741
Show Author Affiliations
Wiktor Franus, Warsaw Univ. of Technology (Poland)
Bartłomiej Twardowski, Warsaw Univ. of Technology (Poland)
Paweł Zawistowski, Warsaw Univ. of Technology (Poland)
Robert M. Nowak, Warsaw Univ. of Technology (Poland)


Published in SPIE Proceedings Vol. 11176:
Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2019
Ryszard S. Romaniuk; Maciej Linczuk, Editor(s)

© SPIE. Terms of Use
Back to Top
PREMIUM CONTENT
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?
close_icon_gray