ISCRAM Digital Library
Home
|
Show All
|
Simple Search
|
Advanced Search
Login
Quick Search:
Field:
main fields
author
title
publication
keywords
abstract
area
conference
contains:
...
1–1 of 1 record found matching your query (
RSS
):
Search & Display Options
Search within Results:
Field:
author
title
year
keywords
abstract
type
publication
abbrev_journal
volume
issue
pages
thesis
publisher
place
editor
series_title
language
area
notes
call_number
serial
contains:
...
Exclude matches
Display Options:
Field:
all fields
keywords & abstract
additional fields
records per page
Select All
Deselect All
<<
1
>>
List View
|
Citations
|
Details
Record
Links
Author
Valentin Barriere
;
Guillaume Jacquet
Title
How does a Pre-Trained Transformer Integrate Contextual Keywords? Application to Humanitarian Computing
Type
Conference Article
Year
2021
Publication
ISCRAM 2021 Conference Proceedings – 18th International Conference on Information Systems for Crisis Response and Management
Abbreviated Journal
Iscram 2021
Volume
Issue
Pages
766-771
Keywords
Transformers, Contextual keywords, Humanitarian Computing, Tweets analysis
Abstract
In a classification task, dealing with text snippets and metadata usually requires to deal with multimodal approaches. When those metadata are textual, it is tempting to use them intrinsically with a pre-trained transformer, in order to leverage the semantic information encoded inside the model. This paper describes how to improve a humanitarian classification task by adding the crisis event type to each tweet to be classified. Based on additional experiments of the model weights and behavior, it identifies how the proposed neural network approach is partially over-fitting the particularities of the Crisis Benchmark, to better highlight how the model is still undoubtedly learning to use and take advantage of the metadata's textual semantics.
Address
European Commission's Joint Research Center; European Commission's Joint Research Center
Corporate Author
Thesis
Publisher
Virginia Tech
Place of Publication
Blacksburg, VA (USA)
Editor
Anouck Adrot; Rob Grace; Kathleen Moore; Christopher W. Zobel
Language
English
Summary Language
English
Original Title
Series Editor
Series Title
Abbreviated Series Title
Series Volume
Series Issue
Edition
ISSN
978-1-949373-61-5
ISBN
Medium
Track
Social Media for Disaster Response and Resilience
Expedition
Conference
18th International Conference on Information Systems for Crisis Response and Management
Notes
valbarrierepro@gmail.com
Approved
no
Call Number
ISCRAM @ idladmin @
Serial
2371
Share this record to Facebook
Select All
Deselect All
<<
1
>>
List View
|
Citations
|
Details
Home
CQL Search
|
Library Search
|
Show Record
|
Extract Citations
Help