Domain adaptation with adversarial training and graph embeddings

Firoj Alam, Shafiq Rayhan Joty, Muhammad Imran

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

The success of deep neural networks (DNNs) is heavily dependent on the availability of labeled data. However, obtaining labeled data is a big challenge in many real-world problems. In such scenarios, a DNN model can leverage labeled and unlabeled data from a related domain, but it has to deal with the shift in data distributions between the source and the target domains. In this paper, we study the problem of classifying social media posts during a crisis event (e.g., Earthquake). For that, we use labeled and unlabeled data from past similar events (e.g., Flood) and unlabeled data for the current event. We propose a novel model that performs adversarial learning based domain adaptation to deal with distribution drifts and graph based semi-supervised learning to leverage unlabeled data within a single unified deep learning framework. Our experiments with two real-world crisis datasets collected from Twitter demonstrate significant improvements over several baselines.

Original languageEnglish
Title of host publicationACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)
PublisherAssociation for Computational Linguistics (ACL)
Pages1077-1087
Number of pages11
ISBN (Electronic)9781948087322
Publication statusPublished - 1 Jan 2018
Event56th Annual Meeting of the Association for Computational Linguistics, ACL 2018 - Melbourne, Australia
Duration: 15 Jul 201820 Jul 2018

Publication series

NameACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)
Volume1

Conference

Conference56th Annual Meeting of the Association for Computational Linguistics, ACL 2018
CountryAustralia
CityMelbourne
Period15/7/1820/7/18

Fingerprint

Supervised learning
Earthquakes
Availability
Experiments
Deep neural networks
Deep learning

ASJC Scopus subject areas

  • Software
  • Computational Theory and Mathematics

Cite this

Alam, F., Rayhan Joty, S., & Imran, M. (2018). Domain adaptation with adversarial training and graph embeddings. In ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers) (pp. 1077-1087). (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers); Vol. 1). Association for Computational Linguistics (ACL).

Domain adaptation with adversarial training and graph embeddings. / Alam, Firoj; Rayhan Joty, Shafiq; Imran, Muhammad.

ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). Association for Computational Linguistics (ACL), 2018. p. 1077-1087 (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers); Vol. 1).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Alam, F, Rayhan Joty, S & Imran, M 2018, Domain adaptation with adversarial training and graph embeddings. in ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers), vol. 1, Association for Computational Linguistics (ACL), pp. 1077-1087, 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018, Melbourne, Australia, 15/7/18.
Alam F, Rayhan Joty S, Imran M. Domain adaptation with adversarial training and graph embeddings. In ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). Association for Computational Linguistics (ACL). 2018. p. 1077-1087. (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)).
Alam, Firoj ; Rayhan Joty, Shafiq ; Imran, Muhammad. / Domain adaptation with adversarial training and graph embeddings. ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). Association for Computational Linguistics (ACL), 2018. pp. 1077-1087 (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)).
@inproceedings{8d01aaac7eb8451fa625afbd6457053a,
title = "Domain adaptation with adversarial training and graph embeddings",
abstract = "The success of deep neural networks (DNNs) is heavily dependent on the availability of labeled data. However, obtaining labeled data is a big challenge in many real-world problems. In such scenarios, a DNN model can leverage labeled and unlabeled data from a related domain, but it has to deal with the shift in data distributions between the source and the target domains. In this paper, we study the problem of classifying social media posts during a crisis event (e.g., Earthquake). For that, we use labeled and unlabeled data from past similar events (e.g., Flood) and unlabeled data for the current event. We propose a novel model that performs adversarial learning based domain adaptation to deal with distribution drifts and graph based semi-supervised learning to leverage unlabeled data within a single unified deep learning framework. Our experiments with two real-world crisis datasets collected from Twitter demonstrate significant improvements over several baselines.",
author = "Firoj Alam and {Rayhan Joty}, Shafiq and Muhammad Imran",
year = "2018",
month = "1",
day = "1",
language = "English",
series = "ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)",
publisher = "Association for Computational Linguistics (ACL)",
pages = "1077--1087",
booktitle = "ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)",

}

TY - GEN

T1 - Domain adaptation with adversarial training and graph embeddings

AU - Alam, Firoj

AU - Rayhan Joty, Shafiq

AU - Imran, Muhammad

PY - 2018/1/1

Y1 - 2018/1/1

N2 - The success of deep neural networks (DNNs) is heavily dependent on the availability of labeled data. However, obtaining labeled data is a big challenge in many real-world problems. In such scenarios, a DNN model can leverage labeled and unlabeled data from a related domain, but it has to deal with the shift in data distributions between the source and the target domains. In this paper, we study the problem of classifying social media posts during a crisis event (e.g., Earthquake). For that, we use labeled and unlabeled data from past similar events (e.g., Flood) and unlabeled data for the current event. We propose a novel model that performs adversarial learning based domain adaptation to deal with distribution drifts and graph based semi-supervised learning to leverage unlabeled data within a single unified deep learning framework. Our experiments with two real-world crisis datasets collected from Twitter demonstrate significant improvements over several baselines.

AB - The success of deep neural networks (DNNs) is heavily dependent on the availability of labeled data. However, obtaining labeled data is a big challenge in many real-world problems. In such scenarios, a DNN model can leverage labeled and unlabeled data from a related domain, but it has to deal with the shift in data distributions between the source and the target domains. In this paper, we study the problem of classifying social media posts during a crisis event (e.g., Earthquake). For that, we use labeled and unlabeled data from past similar events (e.g., Flood) and unlabeled data for the current event. We propose a novel model that performs adversarial learning based domain adaptation to deal with distribution drifts and graph based semi-supervised learning to leverage unlabeled data within a single unified deep learning framework. Our experiments with two real-world crisis datasets collected from Twitter demonstrate significant improvements over several baselines.

UR - http://www.scopus.com/inward/record.url?scp=85063109392&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063109392&partnerID=8YFLogxK

M3 - Conference contribution

T3 - ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)

SP - 1077

EP - 1087

BT - ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)

PB - Association for Computational Linguistics (ACL)

ER -