Distant Supervision from Disparate Sources for Low-Resource\n Part-of-Speech Tagging

2018/08/29 by Barbara Plank, Plank, Barbara, Żeljko Agić +1 · 1 citation
Computer Science · #Natural Language Processing Techniques #Topic Modeling #Multimodal Machine Learning Applications

paper · pdf · doi:10.48550/arxiv.1808.09733

Abstract

We introduce DsDs: a cross-lingual neural part-of-speech tagger that learns\nfrom disparate sources of distant supervision, and realistically scales to\nhundreds of low-resource languages. The model exploits annotation projection,\ninstance selection, tag dictionaries, morphological lexicons, and distributed\nrepresentations, all in a uniform framework. The approach is simple, yet\nsurprisingly effective, resulting in a new state of the art without access to\nany gold annotated data.\n

Cited by

• Unveiling Factors for Enhanced POS Tagging: A Study of Low-Resource Medieval Romance Languages

Related

• The Best of Both Worlds: Lexical Resources To Improve Low-Resource Part-of-Speech Tagging
• Weakly Supervised POS Taggers Perform Poorly on Truly Low-Resource Languages
• Learning when to trust distant supervision: An application to low-resource POS tagging using cross-lingual projection
• Part-of-Speech Tagging with Minimal Lexicalization
• A Grounded Unsupervised Universal Part-of-Speech Tagger for Low-Resource\n Languages