Learning to geolocalise Tweets at a fine-grained level

Jorge David Gonzalez Paule, Yashar Moshfeghi, Craig Macdonald, Iadh Ounis

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

1 Citation (Scopus)
42 Downloads (Pure)

Abstract

Fine-grained geolocation of tweets has become an important feature for reliably performing a wide range of tasks such as real-time event detection, topic detection or disaster and emergency analysis. Recent work adopted a ranking approach to return a predicted location based on content-based similarity to already available individual geotagged tweets. However, this work made use of the IDF weighting model to compute the ranking, which can diminish the quality of the Top-N retrieved tweets. In this work, we adopt a learning to rank approach towards improving the effectiveness of the ranking and increasing the accuracy of fine-grained geolocalisation. To this end we propose a set of features extracted from pairs of geotagged tweets generated within the same fine-grained geographical area (squared areas of size 1 km). Using geotagged tweets from two cities (Chicago and New York, USA), our experimental results show that our learning to rank approach significantly outperforms previous work based on IDF ranking, and improves accuracy of tweet geolocalisation at a fine-grained level.
Original languageEnglish
Title of host publicationCIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management
Place of PublicationNew York
Pages1675-1678
Number of pages4
DOIs
Publication statusPublished - 22 Oct 2018
EventACM International Conference on Information and Knowledge Management - Lingotto, Turin, Italy
Duration: 22 Oct 2018 → 26 Oct 2018
http://www.cikm2018.units.it/#3rdPage

Conference

ConferenceACM International Conference on Information and Knowledge Management
Abbreviated titleCIKM 2018
CountryItaly
CityTurin
Period22/10/18 → 26/10/18
Internet address

Fingerprint

Disasters

Keywords

  • information retrieval
  • Tweet geolocalisation
  • learning to rank

Cite this

Paule, J. D. G., Moshfeghi, Y., Macdonald, C., & Ounis, I. (2018). Learning to geolocalise Tweets at a fine-grained level. In CIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management (pp. 1675-1678). New York. https://doi.org/10.1145/3269206.3269291
Paule, Jorge David Gonzalez ; Moshfeghi, Yashar ; Macdonald, Craig ; Ounis, Iadh. / Learning to geolocalise Tweets at a fine-grained level. CIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management. New York, 2018. pp. 1675-1678
@inproceedings{83e8778c93354a0985ae04f8497812f5,
title = "Learning to geolocalise Tweets at a fine-grained level",
abstract = "Fine-grained geolocation of tweets has become an important feature for reliably performing a wide range of tasks such as real-time event detection, topic detection or disaster and emergency analysis. Recent work adopted a ranking approach to return a predicted location based on content-based similarity to already available individual geotagged tweets. However, this work made use of the IDF weighting model to compute the ranking, which can diminish the quality of the Top-N retrieved tweets. In this work, we adopt a learning to rank approach towards improving the effectiveness of the ranking and increasing the accuracy of fine-grained geolocalisation. To this end we propose a set of features extracted from pairs of geotagged tweets generated within the same fine-grained geographical area (squared areas of size 1 km). Using geotagged tweets from two cities (Chicago and New York, USA), our experimental results show that our learning to rank approach significantly outperforms previous work based on IDF ranking, and improves accuracy of tweet geolocalisation at a fine-grained level.",
keywords = "information retrieval, Tweet geolocalisation, learning to rank",
author = "Paule, {Jorge David Gonzalez} and Yashar Moshfeghi and Craig Macdonald and Iadh Ounis",
year = "2018",
month = "10",
day = "22",
doi = "10.1145/3269206.3269291",
language = "English",
isbn = "9781450360142",
pages = "1675--1678",
booktitle = "CIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management",

}

Paule, JDG, Moshfeghi, Y, Macdonald, C & Ounis, I 2018, Learning to geolocalise Tweets at a fine-grained level. in CIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management. New York, pp. 1675-1678, ACM International Conference on Information and Knowledge Management, Turin, Italy, 22/10/18. https://doi.org/10.1145/3269206.3269291

Learning to geolocalise Tweets at a fine-grained level. / Paule, Jorge David Gonzalez; Moshfeghi, Yashar; Macdonald, Craig; Ounis, Iadh.

CIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management. New York, 2018. p. 1675-1678.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

TY - GEN

T1 - Learning to geolocalise Tweets at a fine-grained level

AU - Paule, Jorge David Gonzalez

AU - Moshfeghi, Yashar

AU - Macdonald, Craig

AU - Ounis, Iadh

PY - 2018/10/22

Y1 - 2018/10/22

N2 - Fine-grained geolocation of tweets has become an important feature for reliably performing a wide range of tasks such as real-time event detection, topic detection or disaster and emergency analysis. Recent work adopted a ranking approach to return a predicted location based on content-based similarity to already available individual geotagged tweets. However, this work made use of the IDF weighting model to compute the ranking, which can diminish the quality of the Top-N retrieved tweets. In this work, we adopt a learning to rank approach towards improving the effectiveness of the ranking and increasing the accuracy of fine-grained geolocalisation. To this end we propose a set of features extracted from pairs of geotagged tweets generated within the same fine-grained geographical area (squared areas of size 1 km). Using geotagged tweets from two cities (Chicago and New York, USA), our experimental results show that our learning to rank approach significantly outperforms previous work based on IDF ranking, and improves accuracy of tweet geolocalisation at a fine-grained level.

AB - Fine-grained geolocation of tweets has become an important feature for reliably performing a wide range of tasks such as real-time event detection, topic detection or disaster and emergency analysis. Recent work adopted a ranking approach to return a predicted location based on content-based similarity to already available individual geotagged tweets. However, this work made use of the IDF weighting model to compute the ranking, which can diminish the quality of the Top-N retrieved tweets. In this work, we adopt a learning to rank approach towards improving the effectiveness of the ranking and increasing the accuracy of fine-grained geolocalisation. To this end we propose a set of features extracted from pairs of geotagged tweets generated within the same fine-grained geographical area (squared areas of size 1 km). Using geotagged tweets from two cities (Chicago and New York, USA), our experimental results show that our learning to rank approach significantly outperforms previous work based on IDF ranking, and improves accuracy of tweet geolocalisation at a fine-grained level.

KW - information retrieval

KW - Tweet geolocalisation

KW - learning to rank

U2 - 10.1145/3269206.3269291

DO - 10.1145/3269206.3269291

M3 - Conference contribution book

SN - 9781450360142

SP - 1675

EP - 1678

BT - CIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management

CY - New York

ER -

Paule JDG, Moshfeghi Y, Macdonald C, Ounis I. Learning to geolocalise Tweets at a fine-grained level. In CIKM '18 Proceedings of the 27th ACM International Conference on Information and Knowledge Management. New York. 2018. p. 1675-1678 https://doi.org/10.1145/3269206.3269291

Access to Document