Content analyses of the international federation of red cross and red crescent societies (ifrc) based on machine learning techniques through twitter
| Autor: | Nazmiye Eligüzel, Cihan Çetinkaya, Türkay Dereli |
|---|---|
| Přispěvatelé: | HKÜ, Mühendislik Fakültesi, Elektrik Elektronik Mühendisliği Bölümü |
| Rok vydání: | 2021 |
| Předmět: |
Atmospheric Science
010504 meteorology & atmospheric sciences LDA Computer science Microblogging 0211 other engineering and technologies Decision tree 02 engineering and technology 01 natural sciences Latent Dirichlet allocation Naive Bayes classifier symbols.namesake Machine learning Earth and Planetary Sciences (miscellaneous) Social media Natural disaster 0105 earth and related environmental sciences Water Science and Technology 021110 strategic defence & security studies Content analyses IFRC Emergency management business.industry Sentiment analysis Data science Topic labeling symbols business |
| Zdroj: | Natural Hazards. 106:2025-2045 |
| ISSN: | 1573-0840 0921-030X |
| DOI: | 10.1007/s11069-021-04527-w |
| Popis: | Intensity of natural disasters has substantially increased; disaster management has gained importance along with this reason. In addition, social media has become an integral part of disaster management. Before, during and after disasters; people use social media and large number of output is obtained through social media activities. In this regard, Twitter is the most popular social media tool as micro blogging. Twitter has also become significant in complex disaster environment for coordinating events. It provides a swift way to collect crowd-sourced information. So, how do humanitarian organizations use Twitter platform? Humanitarian organizations utilize resources and related information while managing disasters. The effective use of social media by humanitarian agencies causes increased peoples’ awareness. The international federation of red cross and Red Crescent Societies (IFRC) is the most significant humanitarian organization that aims providing assistance to people. Thus, the aim of this paper is to analyze IFRC’s activities on Twitter and propose a perspective in the light of theoretical framework. Approximately, 5201 tweets are passed the pre-processing level, some important topics are extracted utilizing word labeling, latent dirichlet allocation (LDA model) and bag of Ngram model and sentiment analysis is applied based on machine learning classification algorithms including Naïve Bayes, support vector machine SVM), decision tree, random forest, neural network and k-nearest neighbor (kNN) classifications. According to the classification accuracies, results demonstrate the superiority of support vector machine among other classification algorithms. This study shows us how IFRC uses Twitter and which topics IFRC emphasizes more. © 2021, The Author(s), under exclusive licence to Springer Nature B.V. part of Springer Nature. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink