{"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/cross-lingual-adaptation-using-structural","title":"Cross-Lingual Adaptation using Structural Correspondence Learning","arxiv_id":"1008.0716","date":"2010-08-04","proceeding":null,"authors":["Peter Prettenhofer","Benno Stein"],"abstract":"Cross-lingual adaptation, a special case of domain adaptation, refers to the\ntransfer of classification knowledge between two languages. In this article we\ndescribe an extension of Structural Correspondence Learning (SCL), a recently\nproposed algorithm for domain adaptation, for cross-lingual adaptation. The\nproposed method uses unlabeled documents from both languages, along with a word\ntranslation oracle, to induce cross-lingual feature correspondences. From these\ncorrespondences a cross-lingual representation is created that enables the\ntransfer of classification knowledge from the source to the target language.\nThe main advantages of this approach over other approaches are its resource\nefficiency and task specificity.\n  We conduct experiments in the area of cross-language topic and sentiment\nclassification involving English as source language and German, French, and\nJapanese as target languages. The results show a significant improvement of the\nproposed method over a machine translation baseline, reducing the relative\nerror due to cross-lingual adaptation by an average of 30% (topic\nclassification) and 59% (sentiment classification). We further report on\nempirical analyses that reveal insights into the use of unlabeled data, the\nsensitivity with respect to important hyperparameters, and the nature of the\ninduced cross-lingual correspondences.","url_abs":"http://arxiv.org/abs/1008.0716v2","url_pdf":"http://arxiv.org/pdf/1008.0716v2.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":"cross-lingual-adaptation-using-structural","repo_url":"https://github.com/pprett/bolt","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null},{"paper_slug":"cross-lingual-adaptation-using-structural","repo_url":"https://github.com/pprett/nut","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"sentiment-analysis","task_name":"Sentiment Analysis"},{"task_slug":"sentiment-classification","task_name":"Sentiment Classification"},{"task_slug":"specificity","task_name":"Specificity"},{"task_slug":"topic-classification","task_name":"Topic Classification"},{"task_slug":"translation","task_name":"Translation"},{"task_slug":"word-translation","task_name":"Word Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1008.0716","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}