Please use this identifier to cite or link to this item:
https://doi.org/10.21256/zhaw-1525
Publication type: | Conference paper |
Type of review: | Not specified |
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 |
DOI: | 10.21256/zhaw-1525 10.1145/3038912.3052611 |
Proceedings: | Proceedings of the 26th International Conference on World Wide Web |
Page(s): | 1045 |
Pages to: | 1052 |
Conference details: | 26th International World Wide Web Conference Committee (IW3C2), Perth, Australia, 3-7 April 2017 |
Issue Date: | 2017 |
Publisher / Ed. Institution: | Association for Computing Machinery |
ISBN: | 9781450349130 |
Language: | English |
Subjects: | Sentiment Analysis |
Subject (DDC): | 006: Special computer methods |
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. |
URI: | https://digitalcollection.zhaw.ch/handle/11475/1851 |
Fulltext version: | Published version |
License (according to publishing contract): | Licence according to publishing contract |
Departement: | School of Engineering |
Organisational Unit: | Institute of Applied Information Technology (InIT) |
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 | Size | Format | |
---|---|---|---|---|
p1045-deriu.pdf | 3.78 MB | Adobe PDF | View/Open |
Show full item record
Deriu, J. M., Lucchi, A., De Luca, V., Severyn, A., Müller, S., Cieliebak, M., Hofmann, T., & Jaggi, M. (2017). Leveraging large amounts of weakly supervised data for multi-language sentiment classification [Conference paper]. Proceedings of the 26th International Conference on World Wide Web, 1045–1052. https://doi.org/10.21256/zhaw-1525
Deriu, J.M. et al. (2017) ‘Leveraging large amounts of weakly supervised data for multi-language sentiment classification’, in Proceedings of the 26th International Conference on World Wide Web. Association for Computing Machinery, pp. 1045–1052. Available at: https://doi.org/10.21256/zhaw-1525.
J. M. Deriu et al., “Leveraging large amounts of weakly supervised data for multi-language sentiment classification,” in Proceedings of the 26th International Conference on World Wide Web, 2017, pp. 1045–1052. doi: 10.21256/zhaw-1525.
DERIU, Jan Milan, Aurelien LUCCHI, Valeria DE LUCA, Aliaksei SEVERYN, Simone MÜLLER, Mark CIELIEBAK, Thomas HOFMANN und Martin JAGGI, 2017. Leveraging large amounts of weakly supervised data for multi-language sentiment classification. In: Proceedings of the 26th International Conference on World Wide Web. Conference paper. Association for Computing Machinery. 2017. S. 1045–1052. ISBN 9781450349130
Deriu, Jan Milan, Aurelien Lucchi, Valeria De Luca, Aliaksei Severyn, Simone Müller, Mark Cieliebak, Thomas Hofmann, and Martin Jaggi. 2017. “Leveraging Large Amounts of Weakly Supervised Data for Multi-Language Sentiment Classification.” Conference paper. In Proceedings of the 26th International Conference on World Wide Web, 1045–52. Association for Computing Machinery. https://doi.org/10.21256/zhaw-1525.
Deriu, Jan Milan, et al. “Leveraging Large Amounts of Weakly Supervised Data for Multi-Language Sentiment Classification.” Proceedings of the 26th International Conference on World Wide Web, Association for Computing Machinery, 2017, pp. 1045–52, https://doi.org/10.21256/zhaw-1525.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.