Introduction

The College of Humanities and Social Sciences (CHSS) will deliver the first international Fake News Detection Contest. Fake news, misinformation, and disinformation have become defining aspects of the information climate in the 21st century. While such phenomena have existed for centuries, the rise of social and digital media has sparked a debate about the relevance of fake news as an emergent cybersecurity threat. Fake news and disinformation has been at the forefront of foreign interference and private sector manipulation of state politics. The revelations about Cambridge Analytica and other actors is most likely the tip of a much larger iceberg.

Like all security threats, fake news and its deployment is constantly evolving, requiring cutting-edge and adaptive solutions. The Digital Humanities team at CHSS is exploring and tackling such phenomena from a holistic and global perspective.

Contest Chairs

Dr. Wajdi Zaghouani : wzaghouani@hbku.edu.qa (Hamad Bin Khalifa University, Qatar)

Dr. Marc Owen Jones: mojones@hbku.edu.qa (Hamad Bin Khalifa University, Qatar)

QICC Awards

1. FAR NLP (AUB) - (15000$)

2. ML Travelers (Sofia) (1000$)

3. UPV (Spain) (5000$)

Dataset Awards

Best Dataset Overall: Team Fatabyyano

Largest Dataset: Team Proppy

Most Innovative Dataset: Team DZ-NLP

Systems Awards

Best Domain Detection: Team Al-Musketeer

Best Bot Detection: Newspeak

Best Fake News Detection: Newspeak

Presentation Awards

Best Poster Presentation : Team Proppy

Best Oral Presentation: Team Truth Seekers

Teams Overall Ranking / Points

1. FAR-NLP 138 Points

2. MLTravellers 137.0 Points

3. UPV 136.0 Points

4. Fatabyyano 134.3 Points

5. Team Proppy 131.7 Points

6. Newspeak 123.1 Points

7. DZ-NLP 120 Points

8. 01Binary 116.7 Points

9. A3-108 105 Points

10. Truth-Seekers 100 Points

11. AI Musketeers 96.9 Points

12. Ali Al-laith 87.1 Points

13. Morroco 71.9 Points

14. ANLP-RG 62.3 Points

15. Fragarach 43.5 Points

16. Misbar 11.0 Points

Task 1A Ranking

1. MLTravellers

2. Newspeak

3. FAR-NLP

4. Team Proppy

5. a3-108

6. UPV

7. Truth-Seekers

8. 01Binary

9. ANLP-RG

10. Ali Al-laith

11. AI Musketeers

12. Morroco

13. Fragarach

14. Fatabyyano

15. DZ-NLP

16. Misbar

Task 1B Ranking

1. FAR-NLP

2. Newspeak

3. 01Binary

4. MLTravellers

5. Team Proppy

6. AI Musketeers

7. a3-108

8. UPV

9. Truth-Seekers

10. Fragarach

11. ANLP-RG

12. Ali Al-laith

13. Morroco

14. DZ-NLP

15. Fatabyyano

16. Misbar

Task 2 Ranking

1. Fatabyyano

2. DZ-NLP

3. UPV

4. Team Proppy

5. a3-108

6. Truth-Seekers

7. Ali Al-laith

8. MLTravellers

9. ANLP-RG

10. Misbar

11. FAR-NLP

12. Newspeak

13. 01Binary

14. AI Musketeers

15. Morroco

16. Fragarach

Contest Tracks

There are 3 tracks in the contest and participants can opt to enroll in 1, 2 or 3 tracks.

Track 1-A : Building a machine learning model to detect fake news and also the news domain. The annotated training corpus will be provided in the upcoming emails.

Track 1-B: Building a machine learning model to detect if a Twitter account is a Bot or Human. The annotated training corpus will be provided in the upcoming emails. The data of this track was provided by Marc Jones.

Track 2 : Collect and annotate a Fake News Corpus as per the guidelines provided in the upcoming emails.

All participating teams will be provided with a common training data set and a common test set. No external manually labelled data sets are allowed. A blind test data set will be used to evaluate the output of the participating teams.

Prizes

$15,000 for the first team, $10,000 for second, and $5,000 for the third.

Invited Speaker

Dr. Preslav Nakov

Principal Scientist at Hamad Bin Khalifa University’s Qatar Computing Research Institute

Important Dates

Application Deadline : April 30 2019

Acceptance Notification : June 20 2019

Training Data Released: September 20 2019

Contest Days : October October 1-2 2019

Awards Announced: October 3 2019

Contest Tentative Schedule

Day 1 – Tuesday, October 1, 2019

08:00 – 08:45 Participant registration

09:00 - 10:10 Opening ceremony

10:10 Coffee break

11:00 – 13:00 Contest activity 1 (Multipurpose Hall 1, 1st floor, 2nd room )

13:00 - 14:30 Participant lunch (Exhibition hall )

14:45 – 16:45 Contest activity 2 (Systems and Dataset Evaluation) / (Multipurpose Hall 1, 1st floor, 2nd room )

Day 2 – Wednesday, October 2, 2019

08:00 - 08:30 Participant check in

08:45 – 10:45 Contest activity 3 (Systems and Dataset Evaluation) / (Multipurpose Hall 1, 1st floor, 2nd room )

11:45 – 13:15 Participant lunch (Exhibition hall )

13:30 – 14:30 Fake News Detection Panel (Conference room 1 )

14:30 – 15:30 Poster Session (Conference room 1 )

16:00 - 17:00 Keynote Talk by Dr. Preslav Nakov :Detecting "Fake News" Before It is Even Written” (Auditorium)

Day 3 – Thursday, October 3, 2019

08:00 - 08:30 Participant check in

09:00 - 11:00 Keynote Talk 2 (Auditorium)

11:15 - 12:00 Keynote Talk 3 (Auditorium)

12:00 - 12:30 Coffee break

12:45 - 13:45 Closing ceremony / Awards Announced

14:00 - 15:30 Participant lunch (Exhibition hall )

16:00 - 17:30 Fake News Students Tutorial (Conference Room 1)

Registration

The general public can attend the event. You need to fill the registration form

If you have any question about the contest, feel free to : wzaghouani@hbku.edu.qa

The full list of the contest participants is available below.