Regions, periods, activities: Uncovering urban dynamics via cross-modal representation learning

Chao Zhang, Keyang Zhang, Quan Yuan, Haoruo Peng, Yu Zheng, Tim Hanratty, Shaowen Wang, Jiawei Han

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

Abstract

With the ever-increasing urbanization process, systematically modeling people’s activities in the urban space is being recognized as a crucial socioeconomic task. This task was nearly impossible years ago due to the lack of reliable data sources, yet the emergence of geo-tagged social media (GTSM) data sheds new light on it. Recently, there have been fruitful studies on discovering geographical topics from GTSM data. However, their high computational costs and strong distributional assumptions about the latent topics hinder them from fully unleashing the power of GTSM. To bridge the gap, we present CrossMap, a novel cross-modal representation learning method that uncovers urban dynamics with massive GTSM data. CrossMap first employs an accelerated mode seeking procedure to detect spatiotemporal hotspots underlying people’s activities. Those detected hotspots not only address spatiotemporal variations, but also largely alleviate the sparsity of the GTSM data. With the detected hotspots, CrossMap then jointly embeds all spatial, temporal, and textual units into the same space using two different strategies: one is reconstruction-based and the other is graph-based. Both strategies capture the correlations among the units by encoding their co-occurrence and neighborhood relationships, and learn low-dimensional representations to preserve such correlations. Our experiments demonstrate that CrossMap not only significantly outperforms state-of-the-art methods for activity recovery and classification, but also achieves much better efficiency.

Original languageEnglish (US)
Title of host publication26th International World Wide Web Conference, WWW 2017
PublisherInternational World Wide Web Conferences Steering Committee
Pages361-370
Number of pages10
ISBN (Print)9781450349130
DOIs
StatePublished - Jan 1 2017
Event26th International World Wide Web Conference, WWW 2017 - Perth, Australia
Duration: Apr 3 2017Apr 7 2017

Publication series

Name26th International World Wide Web Conference, WWW 2017

Other

Other26th International World Wide Web Conference, WWW 2017
CountryAustralia
CityPerth
Period4/3/174/7/17

Fingerprint

Recovery
Costs
Experiments

Keywords

  • Activity
  • Geographical topic
  • Representation learning
  • Social media
  • Spatiotemporal data
  • Twitter
  • Urban dynamics

ASJC Scopus subject areas

  • Software
  • Computer Networks and Communications

Cite this

Zhang, C., Zhang, K., Yuan, Q., Peng, H., Zheng, Y., Hanratty, T., ... Han, J. (2017). Regions, periods, activities: Uncovering urban dynamics via cross-modal representation learning. In 26th International World Wide Web Conference, WWW 2017 (pp. 361-370). [3052601] (26th International World Wide Web Conference, WWW 2017). International World Wide Web Conferences Steering Committee. https://doi.org/10.1145/3038912.3052601

Regions, periods, activities : Uncovering urban dynamics via cross-modal representation learning. / Zhang, Chao; Zhang, Keyang; Yuan, Quan; Peng, Haoruo; Zheng, Yu; Hanratty, Tim; Wang, Shaowen; Han, Jiawei.

26th International World Wide Web Conference, WWW 2017. International World Wide Web Conferences Steering Committee, 2017. p. 361-370 3052601 (26th International World Wide Web Conference, WWW 2017).

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

Zhang, C, Zhang, K, Yuan, Q, Peng, H, Zheng, Y, Hanratty, T, Wang, S & Han, J 2017, Regions, periods, activities: Uncovering urban dynamics via cross-modal representation learning. in 26th International World Wide Web Conference, WWW 2017., 3052601, 26th International World Wide Web Conference, WWW 2017, International World Wide Web Conferences Steering Committee, pp. 361-370, 26th International World Wide Web Conference, WWW 2017, Perth, Australia, 4/3/17. https://doi.org/10.1145/3038912.3052601
Zhang C, Zhang K, Yuan Q, Peng H, Zheng Y, Hanratty T et al. Regions, periods, activities: Uncovering urban dynamics via cross-modal representation learning. In 26th International World Wide Web Conference, WWW 2017. International World Wide Web Conferences Steering Committee. 2017. p. 361-370. 3052601. (26th International World Wide Web Conference, WWW 2017). https://doi.org/10.1145/3038912.3052601
Zhang, Chao ; Zhang, Keyang ; Yuan, Quan ; Peng, Haoruo ; Zheng, Yu ; Hanratty, Tim ; Wang, Shaowen ; Han, Jiawei. / Regions, periods, activities : Uncovering urban dynamics via cross-modal representation learning. 26th International World Wide Web Conference, WWW 2017. International World Wide Web Conferences Steering Committee, 2017. pp. 361-370 (26th International World Wide Web Conference, WWW 2017).
@inproceedings{01ab822283574859bc145d77ed8a1d13,
title = "Regions, periods, activities: Uncovering urban dynamics via cross-modal representation learning",
abstract = "With the ever-increasing urbanization process, systematically modeling people’s activities in the urban space is being recognized as a crucial socioeconomic task. This task was nearly impossible years ago due to the lack of reliable data sources, yet the emergence of geo-tagged social media (GTSM) data sheds new light on it. Recently, there have been fruitful studies on discovering geographical topics from GTSM data. However, their high computational costs and strong distributional assumptions about the latent topics hinder them from fully unleashing the power of GTSM. To bridge the gap, we present CrossMap, a novel cross-modal representation learning method that uncovers urban dynamics with massive GTSM data. CrossMap first employs an accelerated mode seeking procedure to detect spatiotemporal hotspots underlying people’s activities. Those detected hotspots not only address spatiotemporal variations, but also largely alleviate the sparsity of the GTSM data. With the detected hotspots, CrossMap then jointly embeds all spatial, temporal, and textual units into the same space using two different strategies: one is reconstruction-based and the other is graph-based. Both strategies capture the correlations among the units by encoding their co-occurrence and neighborhood relationships, and learn low-dimensional representations to preserve such correlations. Our experiments demonstrate that CrossMap not only significantly outperforms state-of-the-art methods for activity recovery and classification, but also achieves much better efficiency.",
keywords = "Activity, Geographical topic, Representation learning, Social media, Spatiotemporal data, Twitter, Urban dynamics",
author = "Chao Zhang and Keyang Zhang and Quan Yuan and Haoruo Peng and Yu Zheng and Tim Hanratty and Shaowen Wang and Jiawei Han",
year = "2017",
month = "1",
day = "1",
doi = "10.1145/3038912.3052601",
language = "English (US)",
isbn = "9781450349130",
series = "26th International World Wide Web Conference, WWW 2017",
publisher = "International World Wide Web Conferences Steering Committee",
pages = "361--370",
booktitle = "26th International World Wide Web Conference, WWW 2017",

}

TY - GEN

T1 - Regions, periods, activities

T2 - Uncovering urban dynamics via cross-modal representation learning

AU - Zhang, Chao

AU - Zhang, Keyang

AU - Yuan, Quan

AU - Peng, Haoruo

AU - Zheng, Yu

AU - Hanratty, Tim

AU - Wang, Shaowen

AU - Han, Jiawei

PY - 2017/1/1

Y1 - 2017/1/1

N2 - With the ever-increasing urbanization process, systematically modeling people’s activities in the urban space is being recognized as a crucial socioeconomic task. This task was nearly impossible years ago due to the lack of reliable data sources, yet the emergence of geo-tagged social media (GTSM) data sheds new light on it. Recently, there have been fruitful studies on discovering geographical topics from GTSM data. However, their high computational costs and strong distributional assumptions about the latent topics hinder them from fully unleashing the power of GTSM. To bridge the gap, we present CrossMap, a novel cross-modal representation learning method that uncovers urban dynamics with massive GTSM data. CrossMap first employs an accelerated mode seeking procedure to detect spatiotemporal hotspots underlying people’s activities. Those detected hotspots not only address spatiotemporal variations, but also largely alleviate the sparsity of the GTSM data. With the detected hotspots, CrossMap then jointly embeds all spatial, temporal, and textual units into the same space using two different strategies: one is reconstruction-based and the other is graph-based. Both strategies capture the correlations among the units by encoding their co-occurrence and neighborhood relationships, and learn low-dimensional representations to preserve such correlations. Our experiments demonstrate that CrossMap not only significantly outperforms state-of-the-art methods for activity recovery and classification, but also achieves much better efficiency.

AB - With the ever-increasing urbanization process, systematically modeling people’s activities in the urban space is being recognized as a crucial socioeconomic task. This task was nearly impossible years ago due to the lack of reliable data sources, yet the emergence of geo-tagged social media (GTSM) data sheds new light on it. Recently, there have been fruitful studies on discovering geographical topics from GTSM data. However, their high computational costs and strong distributional assumptions about the latent topics hinder them from fully unleashing the power of GTSM. To bridge the gap, we present CrossMap, a novel cross-modal representation learning method that uncovers urban dynamics with massive GTSM data. CrossMap first employs an accelerated mode seeking procedure to detect spatiotemporal hotspots underlying people’s activities. Those detected hotspots not only address spatiotemporal variations, but also largely alleviate the sparsity of the GTSM data. With the detected hotspots, CrossMap then jointly embeds all spatial, temporal, and textual units into the same space using two different strategies: one is reconstruction-based and the other is graph-based. Both strategies capture the correlations among the units by encoding their co-occurrence and neighborhood relationships, and learn low-dimensional representations to preserve such correlations. Our experiments demonstrate that CrossMap not only significantly outperforms state-of-the-art methods for activity recovery and classification, but also achieves much better efficiency.

KW - Activity

KW - Geographical topic

KW - Representation learning

KW - Social media

KW - Spatiotemporal data

KW - Twitter

KW - Urban dynamics

UR - http://www.scopus.com/inward/record.url?scp=85021254271&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021254271&partnerID=8YFLogxK

U2 - 10.1145/3038912.3052601

DO - 10.1145/3038912.3052601

M3 - Conference contribution

AN - SCOPUS:85021254271

SN - 9781450349130

T3 - 26th International World Wide Web Conference, WWW 2017

SP - 361

EP - 370

BT - 26th International World Wide Web Conference, WWW 2017

PB - International World Wide Web Conferences Steering Committee

ER -