Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method

Angelo Basile; Tommaso Caselli

doi:10.1007/978-3-030-58219-7_9

Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method

Angelo Basile, Tommaso Caselli^*

^*Corresponding author for this work

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

4 Citations (Scopus)

Abstract

2019 has been characterized by worldwide waves of protests. Each country’s protests is different but there appear to be common factors. In this paper we present two approaches for identifying protest events in news in English. Our goal is to provide political science and discourse analysis scholars with tools that may facilitate the understanding of this on-going phenomenon. We test our approaches against the ProtestNews Lab 2019 benchmark that challenges systems to perform unsupervised domain adaptation on protest events on three sub-tasks: document classification, sentence classification, and event extraction. Results indicate that developing dedicated architectures and models for each task outperforms simpler solutions based on the propagation of labels from lexical items to documents. Furthermore, we complete the description of our systems with a detailed data analysis to shed light on the limits of the methods.

Original language	English
Title of host publication	Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings
Editors	Avi Arampatzis, Evangelos Kanoulas, Theodora Tsikrika, Stefanos Vrochidis, Hideo Joho, Christina Lioma, Carsten Eickhoff, Aurélie Névéol, Aurélie Névéol, Linda Cappellato, Nicola Ferro
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	97-111
Number of pages	15
ISBN (Print)	9783030582180
DOIs	https://doi.org/10.1007/978-3-030-58219-7_9
Publication status	Published - 2020
Event	11th Conference and Labs of the Evaluation Forum, CLEF 2020 - Thessaloniki, Greece Duration: 22-Sept-2020 → 25-Sept-2020

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12260 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	11th Conference and Labs of the Evaluation Forum, CLEF 2020
Country/Territory	Greece
City	Thessaloniki
Period	22/09/2020 → 25/09/2020

Keywords

Document classification
Event extraction
Protest events
Sentence classification

Access to Document

10.1007/978-3-030-58219-7_9

Handle.net

Persistent link

Cite this

Basile, A., & Caselli, T. (2020). Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method. In A. Arampatzis, E. Kanoulas, T. Tsikrika, S. Vrochidis, H. Joho, C. Lioma, C. Eickhoff, A. Névéol, A. Névéol, L. Cappellato, & N. Ferro (Eds.), Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings (pp. 97-111). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12260 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-58219-7_9

Basile, Angelo ; Caselli, Tommaso. / Protest Event Detection : When Task-Specific Models Outperform an Event-Driven Method. Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings. editor / Avi Arampatzis ; Evangelos Kanoulas ; Theodora Tsikrika ; Stefanos Vrochidis ; Hideo Joho ; Christina Lioma ; Carsten Eickhoff ; Aurélie Névéol ; Aurélie Névéol ; Linda Cappellato ; Nicola Ferro. Springer Science and Business Media Deutschland GmbH, 2020. pp. 97-111 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{be4f7a05a5144b5a88ea99d58154973b,

title = "Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method",

abstract = "2019 has been characterized by worldwide waves of protests. Each country{\textquoteright}s protests is different but there appear to be common factors. In this paper we present two approaches for identifying protest events in news in English. Our goal is to provide political science and discourse analysis scholars with tools that may facilitate the understanding of this on-going phenomenon. We test our approaches against the ProtestNews Lab 2019 benchmark that challenges systems to perform unsupervised domain adaptation on protest events on three sub-tasks: document classification, sentence classification, and event extraction. Results indicate that developing dedicated architectures and models for each task outperforms simpler solutions based on the propagation of labels from lexical items to documents. Furthermore, we complete the description of our systems with a detailed data analysis to shed light on the limits of the methods.",

keywords = "Document classification, Event extraction, Protest events, Sentence classification",

author = "Angelo Basile and Tommaso Caselli",

year = "2020",

doi = "10.1007/978-3-030-58219-7_9",

language = "English",

isbn = "9783030582180",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "97--111",

editor = "Avi Arampatzis and Evangelos Kanoulas and Theodora Tsikrika and Stefanos Vrochidis and Hideo Joho and Christina Lioma and Carsten Eickhoff and Aur{\'e}lie N{\'e}v{\'e}ol and Aur{\'e}lie N{\'e}v{\'e}ol and Linda Cappellato and Nicola Ferro",

booktitle = "Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings",

note = "11th Conference and Labs of the Evaluation Forum, CLEF 2020 ; Conference date: 22-09-2020 Through 25-09-2020",

}

Basile, A & Caselli, T 2020, Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method. in A Arampatzis, E Kanoulas, T Tsikrika, S Vrochidis, H Joho, C Lioma, C Eickhoff, A Névéol, A Névéol, L Cappellato & N Ferro (eds), Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12260 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 97-111, 11th Conference and Labs of the Evaluation Forum, CLEF 2020, Thessaloniki, Greece, 22/09/2020. https://doi.org/10.1007/978-3-030-58219-7_9

Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method. / Basile, Angelo; Caselli, Tommaso.
Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings. ed. / Avi Arampatzis; Evangelos Kanoulas; Theodora Tsikrika; Stefanos Vrochidis; Hideo Joho; Christina Lioma; Carsten Eickhoff; Aurélie Névéol; Aurélie Névéol; Linda Cappellato; Nicola Ferro. Springer Science and Business Media Deutschland GmbH, 2020. p. 97-111 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12260 LNCS).

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

TY - GEN

T1 - Protest Event Detection

T2 - 11th Conference and Labs of the Evaluation Forum, CLEF 2020

AU - Basile, Angelo

AU - Caselli, Tommaso

PY - 2020

Y1 - 2020

N2 - 2019 has been characterized by worldwide waves of protests. Each country’s protests is different but there appear to be common factors. In this paper we present two approaches for identifying protest events in news in English. Our goal is to provide political science and discourse analysis scholars with tools that may facilitate the understanding of this on-going phenomenon. We test our approaches against the ProtestNews Lab 2019 benchmark that challenges systems to perform unsupervised domain adaptation on protest events on three sub-tasks: document classification, sentence classification, and event extraction. Results indicate that developing dedicated architectures and models for each task outperforms simpler solutions based on the propagation of labels from lexical items to documents. Furthermore, we complete the description of our systems with a detailed data analysis to shed light on the limits of the methods.

AB - 2019 has been characterized by worldwide waves of protests. Each country’s protests is different but there appear to be common factors. In this paper we present two approaches for identifying protest events in news in English. Our goal is to provide political science and discourse analysis scholars with tools that may facilitate the understanding of this on-going phenomenon. We test our approaches against the ProtestNews Lab 2019 benchmark that challenges systems to perform unsupervised domain adaptation on protest events on three sub-tasks: document classification, sentence classification, and event extraction. Results indicate that developing dedicated architectures and models for each task outperforms simpler solutions based on the propagation of labels from lexical items to documents. Furthermore, we complete the description of our systems with a detailed data analysis to shed light on the limits of the methods.

KW - Document classification

KW - Event extraction

KW - Protest events

KW - Sentence classification

UR - http://www.scopus.com/inward/record.url?scp=85092161623&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-58219-7_9

DO - 10.1007/978-3-030-58219-7_9

M3 - Conference contribution

AN - SCOPUS:85092161623

SN - 9783030582180

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 97

EP - 111

BT - Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings

A2 - Arampatzis, Avi

A2 - Kanoulas, Evangelos

A2 - Tsikrika, Theodora

A2 - Vrochidis, Stefanos

A2 - Joho, Hideo

A2 - Lioma, Christina

A2 - Eickhoff, Carsten

A2 - Névéol, Aurélie

A2 - Cappellato, Linda

A2 - Ferro, Nicola

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 22 September 2020 through 25 September 2020

ER -

Basile A, Caselli T. Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method. In Arampatzis A, Kanoulas E, Tsikrika T, Vrochidis S, Joho H, Lioma C, Eickhoff C, Névéol A, Névéol A, Cappellato L, Ferro N, editors, Experimental IR Meets Multilinguality, Multimodality, and Interaction - 11th International Conference of the CLEF Association, CLEF 2020, Proceedings. Springer Science and Business Media Deutschland GmbH. 2020. p. 97-111. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-58219-7_9

Protest Event Detection: When Task-Specific Models Outperform an Event-Driven Method

Abstract

Publication series

Conference

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

Cite this