On fine-grained geolocalisation of tweets and real-time traffic incident detection

Jorge David Gonzalez Paule, Yeran Sun, Yashar Moshfeghi

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Recently, geolocalisation of tweets has become important for a wide range of real-time applications, including real-time event detection, topic detection or disaster and emergency analysis. However, the number of relevant geotagged tweets available to enable such tasks remains insufficient. To overcome this limitation, predicting the location of non-geotagged tweets, while challenging, can increase the sample of geotagged data and has consequences for a wide range of applications. In this paper, we propose a location inference method that utilises a ranking approach combined with a majority voting of tweets, where each vote is weighted based on evidence gathered from the ranking. Using geotagged tweets from two cities, Chicago and New York (USA), our experimental results demonstrate that our method (statistically) significantly outperforms state-of-the-art baselines in terms of accuracy and error distance, in both cities, with the cost of decreased coverage. Finally, we investigated the applicability of our method in a real-time scenario by means of a traffic incident detection task. Our analysis shows that our fine-grained geolocalisation method can overcome the limitations of geotagged tweets and precisely map incident-related tweets at the real location of the incident.

LanguageEnglish
JournalInformation Processing and Management
Early online date7 Apr 2018
DOIs
Publication statusE-pub ahead of print - 7 Apr 2018

Fingerprint

incident
traffic
ranking
Disasters
voting
disaster
voter
coverage
scenario
time
Incidents
event
Costs
costs
evidence
Ranking

Keywords

  • fine-grained geolocation
  • information retrieval
  • majority voting
  • traffic incident detection
  • Twitter
  • location-based social networks (LBSN)

Cite this

@article{78df8fb56e0a468d8d5b1eaeb097acc4,
title = "On fine-grained geolocalisation of tweets and real-time traffic incident detection",
abstract = "Recently, geolocalisation of tweets has become important for a wide range of real-time applications, including real-time event detection, topic detection or disaster and emergency analysis. However, the number of relevant geotagged tweets available to enable such tasks remains insufficient. To overcome this limitation, predicting the location of non-geotagged tweets, while challenging, can increase the sample of geotagged data and has consequences for a wide range of applications. In this paper, we propose a location inference method that utilises a ranking approach combined with a majority voting of tweets, where each vote is weighted based on evidence gathered from the ranking. Using geotagged tweets from two cities, Chicago and New York (USA), our experimental results demonstrate that our method (statistically) significantly outperforms state-of-the-art baselines in terms of accuracy and error distance, in both cities, with the cost of decreased coverage. Finally, we investigated the applicability of our method in a real-time scenario by means of a traffic incident detection task. Our analysis shows that our fine-grained geolocalisation method can overcome the limitations of geotagged tweets and precisely map incident-related tweets at the real location of the incident.",
keywords = "fine-grained geolocation, information retrieval, majority voting, traffic incident detection, Twitter, location-based social networks (LBSN)",
author = "Paule, {Jorge David Gonzalez} and Yeran Sun and Yashar Moshfeghi",
year = "2018",
month = "4",
day = "7",
doi = "10.1016/j.ipm.2018.03.011",
language = "English",
journal = "Information Processing and Management",
issn = "0306-4573",

}

On fine-grained geolocalisation of tweets and real-time traffic incident detection. / Paule, Jorge David Gonzalez; Sun, Yeran; Moshfeghi, Yashar.

In: Information Processing and Management, 07.04.2018.

Research output: Contribution to journalArticle

TY - JOUR

T1 - On fine-grained geolocalisation of tweets and real-time traffic incident detection

AU - Paule, Jorge David Gonzalez

AU - Sun, Yeran

AU - Moshfeghi, Yashar

PY - 2018/4/7

Y1 - 2018/4/7

N2 - Recently, geolocalisation of tweets has become important for a wide range of real-time applications, including real-time event detection, topic detection or disaster and emergency analysis. However, the number of relevant geotagged tweets available to enable such tasks remains insufficient. To overcome this limitation, predicting the location of non-geotagged tweets, while challenging, can increase the sample of geotagged data and has consequences for a wide range of applications. In this paper, we propose a location inference method that utilises a ranking approach combined with a majority voting of tweets, where each vote is weighted based on evidence gathered from the ranking. Using geotagged tweets from two cities, Chicago and New York (USA), our experimental results demonstrate that our method (statistically) significantly outperforms state-of-the-art baselines in terms of accuracy and error distance, in both cities, with the cost of decreased coverage. Finally, we investigated the applicability of our method in a real-time scenario by means of a traffic incident detection task. Our analysis shows that our fine-grained geolocalisation method can overcome the limitations of geotagged tweets and precisely map incident-related tweets at the real location of the incident.

AB - Recently, geolocalisation of tweets has become important for a wide range of real-time applications, including real-time event detection, topic detection or disaster and emergency analysis. However, the number of relevant geotagged tweets available to enable such tasks remains insufficient. To overcome this limitation, predicting the location of non-geotagged tweets, while challenging, can increase the sample of geotagged data and has consequences for a wide range of applications. In this paper, we propose a location inference method that utilises a ranking approach combined with a majority voting of tweets, where each vote is weighted based on evidence gathered from the ranking. Using geotagged tweets from two cities, Chicago and New York (USA), our experimental results demonstrate that our method (statistically) significantly outperforms state-of-the-art baselines in terms of accuracy and error distance, in both cities, with the cost of decreased coverage. Finally, we investigated the applicability of our method in a real-time scenario by means of a traffic incident detection task. Our analysis shows that our fine-grained geolocalisation method can overcome the limitations of geotagged tweets and precisely map incident-related tweets at the real location of the incident.

KW - fine-grained geolocation

KW - information retrieval

KW - majority voting

KW - traffic incident detection

KW - Twitter

KW - location-based social networks (LBSN)

UR - https://www.sciencedirect.com/journal/information-processing-and-management

U2 - 10.1016/j.ipm.2018.03.011

DO - 10.1016/j.ipm.2018.03.011

M3 - Article

JO - Information Processing and Management

T2 - Information Processing and Management

JF - Information Processing and Management

SN - 0306-4573

ER -