Within the scope of the project titled “Utilizing Digital Technology for Social Cohesion, Positive Messaging and Peace by Boosting Collaboration, Exchange and Solidarity”, carried out with the financial support of the European Union and the Friedrich Naumann Foundation, in partnership with Sabancı University and Boğaziçi University, the Hrant Dink Foundation organized a session on Computational Social Sciences at the Signal Processing and Communication Applications Congress (SIU).

The results of the 'Hate Speech Detection in Social Media' competition, organized to develop, present and evaluate innovative methods for the detection of hate speech in Turkish texts, were announced. Berrin Yanıkoğlu from Sabancı University made a presentation on the data sets used in the competition, the methodology of the study and the results of the competition.

Of the 20 teams registered for the competition, including participants from Yıldız Technical University, Sabancı University, TOBB University of Economics and Technology and Koç University, eight of them uploaded results for one or more of the sub-problems. The results of the competition, which were determined according to the pre-announced scores, are as follows on a sub-problem basis:

Identification of hate speech category (5 class) on the Israeli-Palestinian conflict
First: Ahmet Tarik Karakas -> Score: 0.5758
Second: Hasan Kemik -> Score: 0.5750
Third: Mustafa Budak, İlknur Durgar Elkahlout, Burcu Şenol, Muhammet Nusret Özateş -> Score: 0.5682

Hate speech binary classification on refugee dataset
First: Kaan Efe Keles and Ömer Kaan Gürbüz -> Score: 0.7687
Second: Hasan Kemik -> Score: 0.7562
Third: Ahmet Tarik Karakas -> Score: 0.7280

Estimation of hate speech intensity in the Israeli-Palestinian conflict
First: Ahmet Tarik Karakas -> Score: 0.3161
Second: Hasan Kemik -> Score: 0.3520
Third: Kaan Efe Keles and Ömer Kaan Gürbüz -> Score: 0.3547

Prediction of hate speech violence on refugee dataset
First: Kaan Efe Keles and Ömer Kaan Gürbüz -> Score: 0.4470
Second: Ahmet Tarik Karakas -> Score: 0.4715
Third: Hasan Kemik -> Score: 0.5448

Congratulations to everyone who participated!
All participants who make at least one successful upload to Kaggle will also receive a certificate.
We would like to thank all the teams who participated in our competition for their interest and effort. We hope it was an enjoyable and instructive experience.

Bu proje Avrupa Birliği tarafından finanse edilmektedir.