Estimation of Twitter user's nationality based on friends and followers information
[ X ]
Tarih
2018
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Pergamon-Elsevier Science Ltd
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Big Data has become very useful in many fields since it provides answers to many important questions that can significantly enhance decision making and process optimization. One of the most interesting domains in big data is the prediction of human features, facts and behaviors. In this paper a new and effective algorithm to predict the nationality of Twitter users is proposed. The proposed algorithm tries to prognosticate the Twitter user's location from their friend's location information only without needing GPS information. Although only approximately 30% of Twitter users write their location information in meaningful form, this paper proves that this percentage is enough to determine the nationality of any Twitter user correctly. The proposed algorithm is applied to estimate the thresholds that will be used to determine the nationality of Twitter users. The results show that our algorithm can correctly classify an average of 90% of the Twitter users. (C) 2017 Elsevier Ltd. All rights reserved.
Açıklama
Al-Abadi, Ahmed Kh. Abbas/0000-0002-4249-3443
Anahtar Kelimeler
Social Media Analyzing, Location Prediction, Twitter User Location, KNIME, Nationality Prediction, Big Data Analyzing
Kaynak
Computers & Electrical Engineering
WoS Q Değeri
Q1
Scopus Q Değeri
Q1
Cilt
66
