Detection, classification and visualization of place-triggerd geotagged tweets

Shinya Hiruta, Takuro Yonezawa, Marko Jurmu, Hideyuki Tokuda

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

11 Citations (Scopus)

Abstract

This paper proposes and evaluates a method to detect and classify tweets that are triggered by places where users locate. Recently, many related works address to detect real world events from social media such as Twitter. However, geotagged tweets often contain noise, which means tweets which are not content-wise related to users' location. This noise is problem for detecting real world events. To address and solve the problem, we define the Place-Triggered Geotagged Tweet, meaning tweets which have both geotag and content-based relation to users' location. We designed and implemented a keyword-based matching technique to detect and classify place-triggered geotagged tweets. We evaluated the performance of our method against a ground truth provided by 18 human classifiers, and achieved 82% accuracy. Additionally, we also present two example applications for visualizing place-triggered geotagged tweets.

Original languageEnglish
Title of host publicationUbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing
Pages956-963
Number of pages8
Publication statusPublished - 2012
Event14th International Conference on Ubiquitous Computing, UbiComp 2012 - Pittsburgh, PA, United States
Duration: 2012 Sep 52012 Sep 8

Other

Other14th International Conference on Ubiquitous Computing, UbiComp 2012
CountryUnited States
CityPittsburgh, PA
Period12/9/512/9/8

Fingerprint

Visualization
Classifiers

Keywords

  • Location-based Services
  • Microblogs
  • Place-triggered Geotagged Tweets

ASJC Scopus subject areas

  • Software

Cite this

Hiruta, S., Yonezawa, T., Jurmu, M., & Tokuda, H. (2012). Detection, classification and visualization of place-triggerd geotagged tweets. In UbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing (pp. 956-963)

Detection, classification and visualization of place-triggerd geotagged tweets. / Hiruta, Shinya; Yonezawa, Takuro; Jurmu, Marko; Tokuda, Hideyuki.

UbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing. 2012. p. 956-963.

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

Hiruta, S, Yonezawa, T, Jurmu, M & Tokuda, H 2012, Detection, classification and visualization of place-triggerd geotagged tweets. in UbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing. pp. 956-963, 14th International Conference on Ubiquitous Computing, UbiComp 2012, Pittsburgh, PA, United States, 12/9/5.
Hiruta S, Yonezawa T, Jurmu M, Tokuda H. Detection, classification and visualization of place-triggerd geotagged tweets. In UbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing. 2012. p. 956-963
Hiruta, Shinya ; Yonezawa, Takuro ; Jurmu, Marko ; Tokuda, Hideyuki. / Detection, classification and visualization of place-triggerd geotagged tweets. UbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing. 2012. pp. 956-963
@inproceedings{104b265e4fa94bf4a8956a5df76e9fee,
title = "Detection, classification and visualization of place-triggerd geotagged tweets",
abstract = "This paper proposes and evaluates a method to detect and classify tweets that are triggered by places where users locate. Recently, many related works address to detect real world events from social media such as Twitter. However, geotagged tweets often contain noise, which means tweets which are not content-wise related to users' location. This noise is problem for detecting real world events. To address and solve the problem, we define the Place-Triggered Geotagged Tweet, meaning tweets which have both geotag and content-based relation to users' location. We designed and implemented a keyword-based matching technique to detect and classify place-triggered geotagged tweets. We evaluated the performance of our method against a ground truth provided by 18 human classifiers, and achieved 82{\%} accuracy. Additionally, we also present two example applications for visualizing place-triggered geotagged tweets.",
keywords = "Location-based Services, Microblogs, Place-triggered Geotagged Tweets",
author = "Shinya Hiruta and Takuro Yonezawa and Marko Jurmu and Hideyuki Tokuda",
year = "2012",
language = "English",
isbn = "9781450312240",
pages = "956--963",
booktitle = "UbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing",

}

TY - GEN

T1 - Detection, classification and visualization of place-triggerd geotagged tweets

AU - Hiruta, Shinya

AU - Yonezawa, Takuro

AU - Jurmu, Marko

AU - Tokuda, Hideyuki

PY - 2012

Y1 - 2012

N2 - This paper proposes and evaluates a method to detect and classify tweets that are triggered by places where users locate. Recently, many related works address to detect real world events from social media such as Twitter. However, geotagged tweets often contain noise, which means tweets which are not content-wise related to users' location. This noise is problem for detecting real world events. To address and solve the problem, we define the Place-Triggered Geotagged Tweet, meaning tweets which have both geotag and content-based relation to users' location. We designed and implemented a keyword-based matching technique to detect and classify place-triggered geotagged tweets. We evaluated the performance of our method against a ground truth provided by 18 human classifiers, and achieved 82% accuracy. Additionally, we also present two example applications for visualizing place-triggered geotagged tweets.

AB - This paper proposes and evaluates a method to detect and classify tweets that are triggered by places where users locate. Recently, many related works address to detect real world events from social media such as Twitter. However, geotagged tweets often contain noise, which means tweets which are not content-wise related to users' location. This noise is problem for detecting real world events. To address and solve the problem, we define the Place-Triggered Geotagged Tweet, meaning tweets which have both geotag and content-based relation to users' location. We designed and implemented a keyword-based matching technique to detect and classify place-triggered geotagged tweets. We evaluated the performance of our method against a ground truth provided by 18 human classifiers, and achieved 82% accuracy. Additionally, we also present two example applications for visualizing place-triggered geotagged tweets.

KW - Location-based Services

KW - Microblogs

KW - Place-triggered Geotagged Tweets

UR - http://www.scopus.com/inward/record.url?scp=84879488347&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84879488347&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84879488347

SN - 9781450312240

SP - 956

EP - 963

BT - UbiComp'12 - Proceedings of the 2012 ACM Conference on Ubiquitous Computing

ER -