RANLP 2017 - Recent Advances in Natural Language Processing Meet Deep Learning 2017
DOI: 10.26615/978-954-452-049-6_031
|View full text |Cite
|
Sign up to set email alerts
|

Graph-based Event Extraction from Twitter

Abstract: Detecting which tweets describe a specific event and clustering them is one of the main challenging tasks related to Social Media currently addressed in the NLP community. Existing approaches have mainly focused on detecting spikes in clusters around specific keywords or Named Entities (NE). However, one of the main drawbacks of such approaches is the difficulty in understanding when the same keywords describe different events. In this paper, we propose a novel approach that exploits NE mentions in tweets and … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
10
0

Year Published

2018
2018
2024
2024

Publication Types

Select...
5
3
2

Relationship

0
10

Authors

Journals

citations
Cited by 22 publications
(10 citation statements)
references
References 22 publications
0
10
0
Order By: Relevance
“…Thus, our model SEDTWik outperforms Twevent in all the three metrics. Edouard et al (2017) and TwitterNews+ (Hasan et al, 2016) also evaluated their models on the same Events2012 dataset (McMinn et al, 2013) but on a different period of tweets and were able to get precision values of 75.0% and 78.0% only. Since we have to manually annotate the results, we did not re-evaluate the results of our model on these tweets but we believe our model would outperform both these models in terms of precision.…”
Section: Event Detection Resultsmentioning
confidence: 99%
“…Thus, our model SEDTWik outperforms Twevent in all the three metrics. Edouard et al (2017) and TwitterNews+ (Hasan et al, 2016) also evaluated their models on the same Events2012 dataset (McMinn et al, 2013) but on a different period of tweets and were able to get precision values of 75.0% and 78.0% only. Since we have to manually annotate the results, we did not re-evaluate the results of our model on these tweets but we believe our model would outperform both these models in terms of precision.…”
Section: Event Detection Resultsmentioning
confidence: 99%
“…The document clusters were generated based on the maximum common subgraphs between each document graph. In a similar attempt, Edouard et al (2017) proposed an event clustering model which leveraged named entities (NE) based directed GoW structure. The GoW was improved by using surrounding context of the graph nodes, NE, to enrich node level information.…”
Section: Related Workmentioning
confidence: 99%
“…Once we can compute similarities, the advantage is that a wide variety of clustering algorithms can be leveraged [21]. For example, community detection algorithms have been used in similar settings [13,31]. One of the most popular algorithms of this type is the Louvain method [10], which relies on modularity-based graph partitioning.…”
Section: Entitymentioning
confidence: 99%