{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/leveraging-large-amounts-of-weakly-supervised","title":"Leveraging Large Amounts of Weakly Supervised Data for Multi-Language Sentiment Classification","arxiv_id":"1703.02504","date":"2017-03-07","proceeding":null,"authors":["Jan Deriu","Aurelien Lucchi","Valeria De Luca","Aliaksei Severyn","Simon Müller","Mark Cieliebak","Thomas Hofmann","Martin Jaggi"],"abstract":"This paper presents a novel approach for multi-lingual sentiment\nclassification in short texts. This is a challenging task as the amount of\ntraining data in languages other than English is very limited. Previously\nproposed multi-lingual approaches typically require to establish a\ncorrespondence to English for which powerful classifiers are already available.\nIn contrast, our method does not require such supervision. We leverage large\namounts of weakly-supervised data in various languages to train a multi-layer\nconvolutional network and demonstrate the importance of using pre-training of\nsuch networks. We thoroughly evaluate our approach on various multi-lingual\ndatasets, including the recent SemEval-2016 sentiment prediction benchmark\n(Task 4), where we achieved state-of-the-art performance. We also compare the\nperformance of our model trained individually for each language to a variant\ntrained for all languages at once. We show that the latter model reaches\nslightly worse - but still acceptable - performance when compared to the single\nlanguage model, while benefiting from better generalization properties across\nlanguages.","url_abs":"http://arxiv.org/abs/1703.02504v1","url_pdf":"http://arxiv.org/pdf/1703.02504v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"leveraging-large-amounts-of-weakly-supervised","repo_url":"https://github.com/spinningbytes/deep-mlsa","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"sentiment-analysis","task_name":"Sentiment Analysis"},{"task_slug":"sentiment-classification","task_name":"Sentiment Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}