Comparing MT approaches for text normalization

Publication type
C1
Publication status
Published
Authors
Matos Veliz, C., De Clercq, O., & Hoste, V.
Series
Proceedings of Recent Advances in Natural Language Processing (RANLP 2019) : natural language processing in a deep learining world
Pagination
740-749
Conference
12th International Conference on 'Recent Advances in Natural Language Processing' (RANLP 2019) (Varna, Bulgaria)
Download
(.pdf)
View in Biblio
(externe link)

Abstract

One of the main characteristics of social media data is the use of non-standard language. Since NLP tools have been trained on traditional text material, their performance drops when applied to social media data. One way to overcome this is to first perform text normalization. In this work, we apply text normalization to noisy English and Dutch text coming from different genres: text messages, message board posts and tweets. We consider the normalization task as a Machine Translation problem and test the two leading paradigms: statistical and neural machine translation. For SMT we explore the added value of varying background corpora for training the language model. For NMT we have a look at data augmentation since the parallel datasets we are working with are limited in size. Our results reveal that when relying on SMT to perform the normalization, it is beneficial to use a background corpus that is close to the genre to be normalized. Regarding NMT, we find that the translations - or normalizations - coming out of this model are far from perfect and that for a low-resource language like Dutch adding additional training data works better than artificially augmenting the data.