Multi-class twitter data categorization and geocoding with a novel computing framework

Abstract

This study details the progress in transportation data analysis with a novel computing framework in keeping with the continuous evolution of the computing technology. The computing framework combines the Labelled Latent Dirichlet Allocation (L-LDA)-incorporated Support Vector Machine (SVM) classifier with the supporting computing strategy on publicly available Twitter data in determining transportation-related events to provide reliable information to travelers. The analytical approach includes analyzing tweets using text classification and geocoding locations based on string similarity. A case study conducted for the New York City and its surrounding areas demonstrates the feasibility of the analytical approach. Approximately 700,010 tweets are analyzed to extract relevant transportation-related information for one week. The SVM classifier achieves more than 85% accuracy in identifying transportation-related tweets from structured data. To further categorize the transportation-related tweets into sub-classes: incident, congestion, construction, special events, and other events, three supervised classifiers are used: L-LDA, SVM, and L-LDA incorporated SVM. Findings from this study demonstrate that the analytical framework, which uses the L-LDA incorporated SVM, can classify roadway transportation-related data from Twitter with over 98.3% accuracy, which is significantly higher than the accuracies achieved by standalone L-LDA and SVM.

City and campus: Exploring the distribution of socio-spatial activities of students of higher education institutes during the global pandemic

“…The Twitter handles were removed from the collected dataset, making the data presented in this paper anonymous. Similar utilization of geotagged Twitter data has a strong precedent in emerging urban literature ( Córdoba et al, 2021 ; García-Palomares et al, 2018 ; Khan et al, 2020 ; Martín et al, 2019 ; Osorio-Arjona & García-Palomares, 2019 ). It must be noted that the study does not advocate that this sample is a homogeneous representation of all students; it might be subject to some biases but it remains within the specified criteria for the study.…”

Combinatorial Approach of Feature Generation for Traffic Event Detection using Social Media Data: CFGA

“…Machine learning methods like support vector machines, logistic regression, and Naive Bayes were applied to study Twitter data for event detection around London [36]. A combination of Support vector machine classifier and Labeled Latent Dirichlet Allocation (L-LDA) was also used to analyze 700,010 tweets and extracted transportation-related information for New York City [37]. In [38], the authors proposed a unified statistical framework that combines the topic modeling based language model and hinge loss Markov random fields (HLMRFs) based model for the road traffic congestion monitoring using Twitter data.…”

Policy Communication on Twitter Social Media: Covid-19 Pandemic Control in Samarinda and Balikpapan City

“…Based on the social media functions that previous research has described, this research will then conduct an in-depth analysis of the activities that occur on each account. Referring to what was conveyed byKhan et al (2020) regarding the opportunity for the presence of social media as an effective medium that can involve any and everyone, this study would provide practical advantages, namely going to contribute to the issue of using social media as a Covid-19 mitigation tool in terms of policy communication, and can give an overview of the extent of communication policies in handling Covid-19 have been developed in two cities, namely Samarinda City and Balikpapan City, through Twitter social media.…”