Social media plays a significant role in disaster management by providing valuable data about affected people, donations and help requests. Recent studies highlight the need to filter information on social media into fine-grained content labels. However, identifying useful information from massive amounts of social media posts during a crisis is a challenging task. In this paper, we propose I-AID, a multimodel approach to automatically categorize tweets into multi-label information types and filter critical information from the enormous volume of social media data. I-AID incorporates three main components: i) a BERT- based encoder to capture the semantics of a tweet and represent as a low-dimensional vector, ii) a graph attention network (GAT) to apprehend correlations between tweets' words/entities and the corresponding information types, and iii) a Relation Network as a learnable distance metric to compute the similarity between tweets and their corresponding information types in a supervised way. We conducted several experiments on two real publicly-available datasets. Our results indicate that I-AID outperforms state-of- the-art approaches in terms of weighted average F1 score by +6% and +4% on the TREC-IS dataset and COVID-19 Tweets, respectively.
%0 Conference Paper
%1 Zahera2021
%A Zahera, Hamada M.
%A Jalota, Rricha
%A Sherif, Mohamed Ahmed
%A Ngonga Ngomo, Axel-Cyrille
%B IEEE Open Access
%D 2021
%J IEEE Open Access
%K daikiri dice eml4u ngonga sherif simba zahera
%T I-AID: Identifying Actionable Information from Disaster-related Tweets
%U https://papers.dice-research.org/2021/IEEEACCESS_I-AID/public.pdf
%X Social media plays a significant role in disaster management by providing valuable data about affected people, donations and help requests. Recent studies highlight the need to filter information on social media into fine-grained content labels. However, identifying useful information from massive amounts of social media posts during a crisis is a challenging task. In this paper, we propose I-AID, a multimodel approach to automatically categorize tweets into multi-label information types and filter critical information from the enormous volume of social media data. I-AID incorporates three main components: i) a BERT- based encoder to capture the semantics of a tweet and represent as a low-dimensional vector, ii) a graph attention network (GAT) to apprehend correlations between tweets' words/entities and the corresponding information types, and iii) a Relation Network as a learnable distance metric to compute the similarity between tweets and their corresponding information types in a supervised way. We conducted several experiments on two real publicly-available datasets. Our results indicate that I-AID outperforms state-of- the-art approaches in terms of weighted average F1 score by +6% and +4% on the TREC-IS dataset and COVID-19 Tweets, respectively.
@inproceedings{Zahera2021,
abstract = {Social media plays a significant role in disaster management by providing valuable data about affected people, donations and help requests. Recent studies highlight the need to filter information on social media into fine-grained content labels. However, identifying useful information from massive amounts of social media posts during a crisis is a challenging task. In this paper, we propose I-AID, a multimodel approach to automatically categorize tweets into multi-label information types and filter critical information from the enormous volume of social media data. I-AID incorporates three main components: i) a BERT- based encoder to capture the semantics of a tweet and represent as a low-dimensional vector, ii) a graph attention network (GAT) to apprehend correlations between tweets' words/entities and the corresponding information types, and iii) a Relation Network as a learnable distance metric to compute the similarity between tweets and their corresponding information types in a supervised way. We conducted several experiments on two real publicly-available datasets. Our results indicate that I-AID outperforms state-of- the-art approaches in terms of weighted average F1 score by +6% and +4% on the TREC-IS dataset and COVID-19 Tweets, respectively.},
added-at = {2021-10-04T14:06:09.000+0200},
author = {Zahera, Hamada M. and Jalota, Rricha and Sherif, Mohamed Ahmed and {Ngonga Ngomo}, Axel-Cyrille},
bdsk-url-1 = {https://papers.dice-research.org/2021/IEEEACCESS_I-AID/public.pdf},
biburl = {https://www.bibsonomy.org/bibtex/25487c4f204a1fe23e4981ef9952daefa/dice-research},
booktitle = {IEEE Open Access},
interhash = {a900e2aa0eff5f03304accbb2dfa9eca},
intrahash = {5487c4f204a1fe23e4981ef9952daefa},
journal = {IEEE Open Access},
keywords = {daikiri dice eml4u ngonga sherif simba zahera},
timestamp = {2023-04-25T16:34:08.000+0200},
title = {I-AID: Identifying Actionable Information from Disaster-related Tweets},
url = {https://papers.dice-research.org/2021/IEEEACCESS_I-AID/public.pdf},
year = 2021
}