ANNOUNCING THE WINNERS OF THE 2024 DATATHON & MACHINE LEARNING COMPETITION:
This year's winner is Team 4 (Team Four against Hate: Daniella Abbott, Barisha Chowdhury, Chanelle Mizrahi, Ben Noyvert). They annotated by far the most tweets and approached the machine learning task with logistic regression. They did a great job in their documentation, showing that they familiarized themselves with the data and the phenomenon they were analyzing. They even chose to visualize some of their results.
Team 3 (Eitan Brochstein, Hunter Herschleb, Maya Nakar, Noah Roth) is awarded second place. They showed great effort in the machine learning challenge: they used gemma2 to classify tweets. Not only did they train a model, they also built a function that automates the annotation process in ISCA's annotation portal, an innovative approach.
Team 1 (Christine Chen, Mikey Mauch, Cohen Schulz, Matthew Widjaja) took third place, using logistic regression to classify tweets.
Special thanks to Trenton Reeves for help in evaluating this year's participant submissions.
Please join us on Sunday, December 8, 2024 at 12PM EST for the online webinar "Online Antisemitism: Trends, Research Gaps, and the Direction of Future Study" with guest speakers Yfat Barak-Cheney, Matthias Becker, and Tal-Or Cohen, where we will honor the winners of this year's competition.
Antisemitism has been widespread on social media for some time. However, it exploded after the horrific Hamas attack in Israel and the ongoing war in Gaza.
The Datathon & Machine Learning Competition is a unique opportunity for high school and college students around the world to get in involved in research at the Institute for the Study of Contemporary Antisemitism (ISCA). Over the course of three weeks, students will participate in a series of workshops on the following topics:
- What is antisemitism? How is it defined? How does it manifest online before and after October 7?
- What is machine learning? How is machine learning helpful when observing and combatting antisemitism online? What challenges exist when using machine learning for this purpose?
- Why is manual annotation a necessary step for automated content detection? How do you manually annotate tweets?
- What computing techniques can be used to analyze online content with machine learning algorithms? How do you build these programs?
After completing three workshops, teams will be tasked with manually annotating a corpus of Tweets about Jewish life and antisemitism. They will then develop their own scripts to identify antisemitic tweets automatically. Teams will be scored on how accurately they identify content as antisemitic, both manually and with the machine learning algorithms that they trained over the course of this competition. For more information, check out last year's competition.
If you are not a high school or undergraduate student and want to participate
This fall, we will be adding an additional track for participants who have already completed their undergraduate degree but still want to participate. If you are a graduate student working on the topic, or if you are just interested in learning more about antisemitism and AI, you can now apply to attend the workshops as well. Participants in this track will not be able to compete for the gift cards. However, they can participate passively or actively and send in their annotations and solutions (scripts) and will receive feedback if they wish to do so.
You can apply as a team of 4-5 people or as an individual and we'll match you. Most people will apply individually.
Prizes:
- Top three teams will be awarded a $500, $300, or $200 VISA / Amazon gift card*
*Type of gift card issued will be at the discretion of event organizers based on regulations in participant's country of residence.
Requirements:
- There is no fee to participate in the competition.
- No prior knowledge is required, but minimal experience with programming will be useful.
- All high school and undergraduate students from all countries are welcome to apply for the competition track. All other interested individuals and groups are welcome to apply for the non-competitive track.
- Participants will need access to a computer and an internet connection. All workshops will be conducted over Zoom.
New Workshop Dates:
- October 6, 2024 @ 12-2 pm ET: Online Antisemitism Before and After October 7.
- October 13, 2024 @ 12-2 pm ET: Manual Annotation: Why and How?
- October 20, 2024 @ 12-2 pm ET: Automated Content Detection: The Basics
Resources and Session Recordings:
IHRA Definition of Antisemitism
Workshop 1: Online Antisemitism Before and After October 7
Workshop 2: Mannual Annotation: Why and How?
Workshop 3: Antisemitism Detection in Social Media
Contact: Project Manager: Katharina Soemer ksoemer@iu.edu
Funded by:
Diane M. Druck
Academic Engagement Network
Borns Jewish Studies Program
IU Office of the Vice President for Diversity, Equity, and Inclusion
Ostrom Workshop - Program on Cybersecurity and Internet Governance
Supported by:
IU Data Science Club
IU Hillel
Chabad of Bloomington
Natural Language Processing Lab
Faculty Assistance in Data Science (FADS)
IU Research
The College of Arts