LDA topic modeling on twitter data concerning immigrants and refugees

Ergül, Halil İbrahim and Terzioğlu, Ayşecan and Tercan, Murat and Yanıkoğlu, Berrin and Arın, İnanç (2023) LDA topic modeling on twitter data concerning immigrants and refugees. In: 31st Signal Processing and Communications Applications Conference (SIU), Istanbul, Turkiye

[thumbnail of SIU_LDA.pdf] PDF
Restricted to Registered users only

Download (653kB) | Request a copy


In this study, the attitudes and opinions of Twitter users in Turkey towards immigrants have been examined to see how people express their thoughts and opinions about immigrants in Turkey and whether there are any dominant and interpretable topics that emerge. After a comprehensive pre-preprocessing, latent themes in the tweets are discovered using the Latent Dirichlet Allocation (LDA) topic modeling methodology. As the result of this analysis, 14 topics have emerged as meaningful and interpretable. The study is done over a small dataset and is somewhat limited; however, the results can shed light on the perspectives of Twitter users towards immigrants and refugees.
Item Type: Papers in Conference Proceedings
Uncontrolled Keywords: anti-immigrant; computational social sciences; hate speech; natural language processing; topic modeling
Subjects: H Social Sciences > H Social Sciences (General)
Divisions: Faculty of Arts and Social Sciences > Academic programs > Cultural Studies
Faculty of Arts and Social Sciences
Depositing User: Ayşecan Terzioğlu
Date Deposited: 03 Oct 2023 12:50
Last Modified: 07 Feb 2024 10:46
URI: https://research.sabanciuniv.edu/id/eprint/48028

Actions (login required)

View Item
View Item