Multi-task learning for spatio-temporal event forecasting

Liang Zhao, Qian Sun, Jieping Ye, Feng Chen, Chang Tien Lu, Naren Ramakrishnan

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

52 Citations (Scopus)

Abstract

Spatial event forecasting from social media is an important problem but encounters critical challenges, such as dynamic patterns of features (keywords) and geographic heterogeneity (e.g., spatial correlations, imbalanced samples, and different populations in different locations). Most existing approaches (e.g., LASSO regression, dynamic query expansion, and burst detection) are designed to address some of these challenges, but not all of them. This paper proposes a novel multi-task learning framework which aims to concurrently address all the challenges. Specifically, given a collection of locations (e.g., cities), we propose to build forecasting models for all locations simultaneously by extracting and utilizing appropriate shared information that effectively increases the sample size for each location, thus improving the forecasting performance. We combine both static features derived from a predefined vocabulary by domain experts and dynamic features generated from dynamic query expansion in a multi-task feature learning framework; we investigate different strategies to balance homogeneity and diversity between static and dynamic terms. Efficient algorithms based on Iterative Group Hard Thresholding are developed to achieve efficient and effective model training and prediction. Extensive experimental evaluations on Twitter data from four different countries in Latin America demonstrated the effectiveness of our proposed approach.

Original languageEnglish (US)
Title of host publicationProceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
PublisherAssociation for Computing Machinery
Pages1503-1512
Number of pages10
Volume2015-August
ISBN (Print)9781450336642
DOIs
StatePublished - Aug 10 2015
Externally publishedYes
Event21st ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2015 - Sydney, Australia
Duration: Aug 10 2015Aug 13 2015

Other

Other21st ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2015
CountryAustralia
CitySydney
Period8/10/158/13/15

Keywords

  • Dynamic query expansion
  • Event forecasting
  • Hard thresholding
  • LASSO
  • Multi-task learning

ASJC Scopus subject areas

  • Software
  • Information Systems

Cite this

Zhao, L., Sun, Q., Ye, J., Chen, F., Lu, C. T., & Ramakrishnan, N. (2015). Multi-task learning for spatio-temporal event forecasting. In Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (Vol. 2015-August, pp. 1503-1512). Association for Computing Machinery. https://doi.org/10.1145/2783258.2783377

Multi-task learning for spatio-temporal event forecasting. / Zhao, Liang; Sun, Qian; Ye, Jieping; Chen, Feng; Lu, Chang Tien; Ramakrishnan, Naren.

Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. 2015-August Association for Computing Machinery, 2015. p. 1503-1512.

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

Zhao, L, Sun, Q, Ye, J, Chen, F, Lu, CT & Ramakrishnan, N 2015, Multi-task learning for spatio-temporal event forecasting. in Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. vol. 2015-August, Association for Computing Machinery, pp. 1503-1512, 21st ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2015, Sydney, Australia, 8/10/15. https://doi.org/10.1145/2783258.2783377
Zhao L, Sun Q, Ye J, Chen F, Lu CT, Ramakrishnan N. Multi-task learning for spatio-temporal event forecasting. In Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. 2015-August. Association for Computing Machinery. 2015. p. 1503-1512 https://doi.org/10.1145/2783258.2783377
Zhao, Liang ; Sun, Qian ; Ye, Jieping ; Chen, Feng ; Lu, Chang Tien ; Ramakrishnan, Naren. / Multi-task learning for spatio-temporal event forecasting. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. 2015-August Association for Computing Machinery, 2015. pp. 1503-1512
@inproceedings{c0f0fc2ed1234cc1baabd226385225a4,
title = "Multi-task learning for spatio-temporal event forecasting",
abstract = "Spatial event forecasting from social media is an important problem but encounters critical challenges, such as dynamic patterns of features (keywords) and geographic heterogeneity (e.g., spatial correlations, imbalanced samples, and different populations in different locations). Most existing approaches (e.g., LASSO regression, dynamic query expansion, and burst detection) are designed to address some of these challenges, but not all of them. This paper proposes a novel multi-task learning framework which aims to concurrently address all the challenges. Specifically, given a collection of locations (e.g., cities), we propose to build forecasting models for all locations simultaneously by extracting and utilizing appropriate shared information that effectively increases the sample size for each location, thus improving the forecasting performance. We combine both static features derived from a predefined vocabulary by domain experts and dynamic features generated from dynamic query expansion in a multi-task feature learning framework; we investigate different strategies to balance homogeneity and diversity between static and dynamic terms. Efficient algorithms based on Iterative Group Hard Thresholding are developed to achieve efficient and effective model training and prediction. Extensive experimental evaluations on Twitter data from four different countries in Latin America demonstrated the effectiveness of our proposed approach.",
keywords = "Dynamic query expansion, Event forecasting, Hard thresholding, LASSO, Multi-task learning",
author = "Liang Zhao and Qian Sun and Jieping Ye and Feng Chen and Lu, {Chang Tien} and Naren Ramakrishnan",
year = "2015",
month = "8",
day = "10",
doi = "10.1145/2783258.2783377",
language = "English (US)",
isbn = "9781450336642",
volume = "2015-August",
pages = "1503--1512",
booktitle = "Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining",
publisher = "Association for Computing Machinery",

}

TY - GEN

T1 - Multi-task learning for spatio-temporal event forecasting

AU - Zhao, Liang

AU - Sun, Qian

AU - Ye, Jieping

AU - Chen, Feng

AU - Lu, Chang Tien

AU - Ramakrishnan, Naren

PY - 2015/8/10

Y1 - 2015/8/10

N2 - Spatial event forecasting from social media is an important problem but encounters critical challenges, such as dynamic patterns of features (keywords) and geographic heterogeneity (e.g., spatial correlations, imbalanced samples, and different populations in different locations). Most existing approaches (e.g., LASSO regression, dynamic query expansion, and burst detection) are designed to address some of these challenges, but not all of them. This paper proposes a novel multi-task learning framework which aims to concurrently address all the challenges. Specifically, given a collection of locations (e.g., cities), we propose to build forecasting models for all locations simultaneously by extracting and utilizing appropriate shared information that effectively increases the sample size for each location, thus improving the forecasting performance. We combine both static features derived from a predefined vocabulary by domain experts and dynamic features generated from dynamic query expansion in a multi-task feature learning framework; we investigate different strategies to balance homogeneity and diversity between static and dynamic terms. Efficient algorithms based on Iterative Group Hard Thresholding are developed to achieve efficient and effective model training and prediction. Extensive experimental evaluations on Twitter data from four different countries in Latin America demonstrated the effectiveness of our proposed approach.

AB - Spatial event forecasting from social media is an important problem but encounters critical challenges, such as dynamic patterns of features (keywords) and geographic heterogeneity (e.g., spatial correlations, imbalanced samples, and different populations in different locations). Most existing approaches (e.g., LASSO regression, dynamic query expansion, and burst detection) are designed to address some of these challenges, but not all of them. This paper proposes a novel multi-task learning framework which aims to concurrently address all the challenges. Specifically, given a collection of locations (e.g., cities), we propose to build forecasting models for all locations simultaneously by extracting and utilizing appropriate shared information that effectively increases the sample size for each location, thus improving the forecasting performance. We combine both static features derived from a predefined vocabulary by domain experts and dynamic features generated from dynamic query expansion in a multi-task feature learning framework; we investigate different strategies to balance homogeneity and diversity between static and dynamic terms. Efficient algorithms based on Iterative Group Hard Thresholding are developed to achieve efficient and effective model training and prediction. Extensive experimental evaluations on Twitter data from four different countries in Latin America demonstrated the effectiveness of our proposed approach.

KW - Dynamic query expansion

KW - Event forecasting

KW - Hard thresholding

KW - LASSO

KW - Multi-task learning

UR - http://www.scopus.com/inward/record.url?scp=84954148765&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84954148765&partnerID=8YFLogxK

U2 - 10.1145/2783258.2783377

DO - 10.1145/2783258.2783377

M3 - Conference contribution

SN - 9781450336642

VL - 2015-August

SP - 1503

EP - 1512

BT - Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

PB - Association for Computing Machinery

ER -