Towards real-time and unsupervised campaign detection in social media

Dennis Assenmacher; Lena Adam; Heike Trautmann; Christian Grimme

Towards real-time and unsupervised campaign detection in social media

Dennis Assenmacher
, Lena Adam
, Heike Trautmann
, Christian Grimme

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

6 Citations (Scopus)

21 Downloads (Pure)

Abstract

The detection of orchestrated and potentially manipulative campaigns in social media is far more meaningful than analyzing single account behaviour but also more challenging in terms of pattern recognition, data processing, and computational complexity. While supervised learning methods need an enormous amount of reliable ground truth data to find rather inflexible patterns, classical unsupervised learning techniques need a lot of computational power to handle large amount of data. This makes them infeasible for realtime analysis. In this work, we demonstrate the applicability of text stream clustering for the real-time detection of coordinated campaigns.

Original language	English
Title of host publication	Proceedings of the 33rd International Florida Artificial Intelligence Research Society Conference, FLAIRS 2020
Editors	Eric Bell, Roman Bartak
Publisher	AAAI
Pages	303-306
Number of pages	4
ISBN (Electronic)	9781577358213
Publication status	Published - 2020
Externally published	Yes
Event	33rd International FLAIRS Conference, FLAIRS 2020 - Virtual Event, United States Duration: 17 May 2020 → 20 May 2020 Conference number: 33

Conference

Conference	33rd International FLAIRS Conference, FLAIRS 2020
Abbreviated title	FLAIRS 2020
Country/Territory	United States
City	Virtual Event
Period	17/05/20 → 20/05/20

Access to Document

ArticleFinal published version, 564 KB

https://aaai.org/papers/303-flairs-2020-18452/

Cite this

@inproceedings{4d3e5bd60fe947008fdb9a68903af85d,

title = "Towards real-time and unsupervised campaign detection in social media",

abstract = "The detection of orchestrated and potentially manipulative campaigns in social media is far more meaningful than analyzing single account behaviour but also more challenging in terms of pattern recognition, data processing, and computational complexity. While supervised learning methods need an enormous amount of reliable ground truth data to find rather inflexible patterns, classical unsupervised learning techniques need a lot of computational power to handle large amount of data. This makes them infeasible for realtime analysis. In this work, we demonstrate the applicability of text stream clustering for the real-time detection of coordinated campaigns.",

author = "Dennis Assenmacher and Lena Adam and Heike Trautmann and Christian Grimme",

year = "2020",

language = "English",

pages = "303--306",

editor = "Eric Bell and Roman Bartak",

booktitle = "Proceedings of the 33rd International Florida Artificial Intelligence Research Society Conference, FLAIRS 2020",

publisher = "AAAI",

address = "United States",

}

Assenmacher, D, Adam, L, Trautmann, H & Grimme, C 2020, Towards real-time and unsupervised campaign detection in social media. in E Bell & R Bartak (eds), Proceedings of the 33rd International Florida Artificial Intelligence Research Society Conference, FLAIRS 2020. AAAI, pp. 303-306, 33rd International FLAIRS Conference, FLAIRS 2020, Virtual Event, United States, 17/05/20. <https://aaai.org/papers/303-flairs-2020-18452/>

Towards real-time and unsupervised campaign detection in social media. / Assenmacher, Dennis; Adam, Lena; Trautmann, Heike et al.
Proceedings of the 33rd International Florida Artificial Intelligence Research Society Conference, FLAIRS 2020. ed. / Eric Bell; Roman Bartak. AAAI, 2020. p. 303-306.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Towards real-time and unsupervised campaign detection in social media

AU - Assenmacher, Dennis

AU - Adam, Lena

AU - Trautmann, Heike

AU - Grimme, Christian

N1 - Conference code: 33

PY - 2020

Y1 - 2020

N2 - The detection of orchestrated and potentially manipulative campaigns in social media is far more meaningful than analyzing single account behaviour but also more challenging in terms of pattern recognition, data processing, and computational complexity. While supervised learning methods need an enormous amount of reliable ground truth data to find rather inflexible patterns, classical unsupervised learning techniques need a lot of computational power to handle large amount of data. This makes them infeasible for realtime analysis. In this work, we demonstrate the applicability of text stream clustering for the real-time detection of coordinated campaigns.

AB - The detection of orchestrated and potentially manipulative campaigns in social media is far more meaningful than analyzing single account behaviour but also more challenging in terms of pattern recognition, data processing, and computational complexity. While supervised learning methods need an enormous amount of reliable ground truth data to find rather inflexible patterns, classical unsupervised learning techniques need a lot of computational power to handle large amount of data. This makes them infeasible for realtime analysis. In this work, we demonstrate the applicability of text stream clustering for the real-time detection of coordinated campaigns.

UR - https://www.scopus.com/pages/publications/85090088298

M3 - Conference contribution

SP - 303

EP - 306

BT - Proceedings of the 33rd International Florida Artificial Intelligence Research Society Conference, FLAIRS 2020

A2 - Bell, Eric

A2 - Bartak, Roman

PB - AAAI

T2 - 33rd International FLAIRS Conference, FLAIRS 2020

Y2 - 17 May 2020 through 20 May 2020

ER -

Towards real-time and unsupervised campaign detection in social media

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this