Syntactic parsing for Urdu improves with shared learning and word context
Multi-Task Learning by using Contextualized Word Representations for Syntactic Parsing of a Morphologically Rich Language
Computation and LanguageArtificial Intelligence
Summary
Parsing sentences to understand their structure is hard for Urdu because its words change form a lot. The authors turned an Urdu grammar treebank into a different format, created a new way to label sentence parts, and trained word representations from lots of Urdu text. They taught their system to do two parsing tasks at once, which improved its accuracy. This approach leads to better understanding of Urdu sentence structures than before.
What this means in practice
- •For natural language processing engineers: Improve Urdu language parsing components in NLP pipelines by using multi-task learning models with contextualized word embeddings.
- •For machine translation developers: Enhance translation quality for Urdu by integrating better syntactic parsing tools trained on large web corpora.
Authors
Toqeer Ehsan, Miriam Butt, Sarmad Hussain, Hassan Alhuzali, Ali Al-Laith
Abstract
We address the challenge of syntactic parsing for Urdu, a morphologically rich language, and present state-of-the-art results for both constituency and dependency parsing. This paper offers four major contributions: 1) the conversion of the CLE-UTB phrase structure treebank into a dependency treebank by developing language-specific head-word and phrase-to-dependency label mapping rules; 2) a novel sequence labeling scheme that transforms the parsing task into a unified representation; 3) the training of contextualized word representations on a large 220 million tokens Urdu corpus collected from the web; and 4) development of parsing framework using two learning paradigms, single-task and multi-task learning. Several post-processing rules are applied to improve the quality of the automatically converted dependency structure treebank. The proposed sequence labeling scheme enables the use of a shared architecture that learns the syntactic structures from both grammatical structures simultaneously and hence improves generalization. Experiments show that the multi-task learning setup significantly enhances parsing performance, achieving an F1 score of 91.39 for constituency parsing (an improvement of 3.29 points) and a labeled attachment score of 85.69 for dependency parsing (an improvement of 1.49 points). These results demonstrate that learning cross-task representations provides measurable benefits and advances the state of syntactic parsing for Urdu.