Learning how to active learn by dreaming

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

14 Citations (Scopus)


Heuristic-based active learning (AL) methods are limited when the data distribution of the underlying learning problems vary. Recent data-driven AL policy learning methods are also restricted to learn from closely related domains. We introduce a new sample-efficient method that learns the AL policy directly on the target domain of interest by using wake and dream cycles. Our approach interleaves between querying the annotation of the selected datapoints to update the underlying student learner and improving AL policy using simulation where the current student learner acts as an imperfect annotator. We evaluate our method on cross-domain and cross-lingual text classification and named entity recognition tasks. Experimental results show that our dream-based AL policy training strategy is more effective than applying the pretrained policy without further fine-tuning, and better than the existing strong baseline methods that use heuristics or reinforcement learning.

Original languageEnglish
Title of host publicationProceedings of the 57th Annual Meeting of the Association for Computational Linguistics
EditorsAnna Korhonen, David Traum, Lluís Màrquez
Place of PublicationFlorence Italy
PublisherAssociation for Computational Linguistics (ACL)
Number of pages11
ISBN (Electronic)9781950737482
Publication statusPublished - Jul 2020
EventAnnual Meeting of the Association of Computational Linguistics 2019 - Florence, Italy
Duration: 28 Jul 20192 Aug 2019
Conference number: 57th
https://www.aclweb.org/anthology/events/acl-2019/ (Proceedings)


ConferenceAnnual Meeting of the Association of Computational Linguistics 2019
Abbreviated titleACL 2019
Internet address

Cite this