Comparison of different machine learning techniques on location extraction by utilizing geo-tagged tweets: A case study

Abstract

In emergencies, Twitter is an important platform to get situational awareness simultaneously. Therefore, information about Twitter users' location is a fundamental aspect to understand the disaster effects. But location extraction is a challenging task. Most of the Twitter users do not share their locations in their tweets. In that respect, there are different methods proposed for location extraction which cover different fields such as statistics, machine learning, etc. This study is a sample study that utilizes geo-tagged tweets to demonstrate the importance of the location in disaster management by taking three cases into consideration. In our study, tweets are obtained by utilizing the "earthquake" keyword to determine the location of Twitter users. Tweets are evaluated by utilizing the Latent Dirichlet Allocation (LDA) topic model and sentiment analysis through machine learning classification algorithms including the Multinomial and Gaussian Naïve Bayes, Support Vector Machine (SVM), Decision Tree, Random Forest, Extra Trees, Neural Network, k Nearest Neighbor (kNN), Stochastic Gradient Descent (SGD), and Adaptive Boosting (AdaBoost) classifications. Therefore, 10 different machine learning algorithms are applied in our study by utilizing sentiment analysis based on location-specific disasterrelated tweets by aiming fast and correct response in a disaster situation. In addition, the effectiveness of each algorithm is evaluated in order to gather the right machine learning algorithm. Moreover, topic extraction via LDA is provided to comprehend the situation after a disaster. The gathered results from the application of three cases indicate that Multinomial Naïve Bayes and Extra Trees machine learning algorithms give the best results with an F-measure value over 80%. The study aims to provide a quick response to earthquakes by applying the aforementioned techniques.

Supply chain risk identification: a real-time data-mining approach

“…Twitter). Though studies in the literature applied Twitter data (Su and Chen, 2018 for supplier selection; Eligüzel et al., 2020 for disaster management), they focus only on a particular risk factor. By contrast, this study includes multiple viewpoints on SC risks to identify the threats.…”

“…According to the study by Chae (2015), a Twitter analytics framework helped to analyze the polarity of tweets using sentiment analysis from SC experts related to the organization's sales performance and disruption. Concerning disaster management, Eligüzel et al. (2020) analyzed different ML algorithms to provide a quick response to earth-quake from the tweets based on location.…”

Identifying and Characterizing the Propagation Scale of COVID-19 Situational Information on Twitter: A Hybrid Text Analytic Approach

“…In a case study, Catherine et al investigate the usage analysis of the Twitter handle of the Mayor of Houston during the hurricanes Sandy and Harvey in August and September of 2017 [29]. In another study, researchers compared machine learning algorithms in the extraction of geo-tagged tweets during emergencies; they used 10 machine learning classifiers on location-oriented disaster-related tweets [30]. A seasonal influenza surveillance system was built on the basis of topics extracted from Twitter [31].…”

Accuracy of a pre-trained sentiment analysis (SA) classification model on tweets related to emergency response and early recovery assessment: the case of 2019 Albanian earthquake

“…These social sensors have the potential to speed up the procedures for intensity calculations (Kropivnitskaya et al 2017a, b). Instead, Twitter is a platform to get situational awareness (Eligüzel et al 2020) during the emergency or relief phase (Contreras 2016) after a disaster. A correlation between the number of tweets and the intensity of an earthquake was observed for the first time in 2010 during the Tohoku earthquake.…”