Please use this identifier to cite or link to this item:
Title: Leveraging large amounts of weakly supervised data for multi-language sentiment classification
Authors : Deriu, Jan Milan
Lucchi, Aurelien
De Luca, Valeria
Severyn, Aliaksei
Müller, Simone
Cieliebak, Mark
Hofmann, Thomas
Jaggi, Martin
Proceedings: Proceedings of the 26th International Conference on World Wide Web
Pages : 1045
Pages to: 1052
Conference details: 26th International World Wide Web Conference Committee (IW3C2), Perth, Australia, April 3–7, 2017
Publisher / Ed. Institution : ACM Press
Issue Date: Apr-2017
License (according to publishing contract) : Licence according to publishing contract
Type of review: Not specified
Language : English
Subjects : Sentiment Analysis
Subject (DDC) : 004: Computer science
005: Computer programming, programs and data
Abstract: This paper presents a novel approach for multi-lingual sentiment classification in short texts. This is a challenging task as the amount of training data in languages other than English is very limited. Previously proposed multi-lingual approaches typically require to establish a correspondence to English for which powerful classifiers are already available. In contrast, our method does not require such supervision. We leverage large amounts of weakly-supervised data in various languages to train a multi-layer convolutional network and demonstrate the importance of using pre-training of such networks. We thoroughly evaluate our approach on various multi-lingual datasets, including the recent SemEval-2016 sentiment prediction benchmark (Task 4), where we achieved state-of-the-art performance. We also compare the performance of our model trained individually for each language to a variant trained for all languages at once. We show that the latter model reaches slightly worse - but still acceptable - performance when compared to the single language model, while benefiting from better generalization properties across languages.
Departement: School of Engineering
Organisational Unit: Institute of Applied Information Technology (InIT)
Publication type: Conference Paper
DOI : 10.1145/3038912.3052611
ISBN: 9781450349130
Published as part of the ZHAW project : DeepText: Intelligente Textanalyse mit Deep Learning
Appears in Collections:Publikationen School of Engineering

Files in This Item:
File Description SizeFormat 
p1045-deriu.pdf3.78 MBAdobe PDFThumbnail

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.