h ckAIhon
The Artificial Intelligence
Club Presents
l
4PM
L2012
make machines
Building L, Main
smarter & us
21/4 cooler :P
BETTER COMMUNITIES
** Cash Prizes and Participation Ceritificates Awarded
Categories
01 Voices of the Community -
Natural Language Processing
02 Vision for All -
Computer Vision
Community Decisions -
03 Data Analytics
Spring 2025
Hackathon Booklet
h ckAIhon
l
HackAIthon 1st Edition By Artificial Intelligence Club
Dear Members,
Yes, it’s another hackathon, but we promise it’ll be worth the extra coffee ;)
We are honored to welcome you to the very 1st edition of HackAIthon, the door to
innovation and the perfect opportunity to dive into the world of AI.
Choose from NLP to analyze community opinions and emotions, Computer Vision to
make life more inclusive, or Tabular Data to make smarter decisions. Each track is a
chance to show us how AI can improve our everyday lives and build a better tomorrow.
This hackathon is built for everyone, from beginners to those who casually fine-tune
LLaMA models in their free time— we see you :)
We’re not just looking for the most accurate models, we want creativity, impact, and
solutions that make people say, “Whoa, I never thought of that!” So, grab your laptop,
some snacks, and let’s build something awesome.
AI for good, AI for fun— we’re hyped to have you here. Let’s make this first edition one
to remember!
Yours Sincerely,
AI Club Team
Hackathon Timeline
01 8th April Registration Opens
02 15th April Orientation and Q&A Session (virtual)
03 17th April Registration Closes
04 21st April Presentations and Winners Announced
Page | 1
h ckAIhon
l
01 - Voices of the Community By Artificial Intelligence Club
Natural Language Processing Task
Participants will develop a model that can analyze public opinions from social media,
forums, or community feedback datasets and extract key concerns, emotions, or
topics to help local authorities and organizations understand their community better.
You may focus on a particular community issue (e.g., public transport feedback,
campus life, mental health posts).
Judging Criteria
Innovation & Relevance Creativity of feature extraction and relevance to community needs. (30%)
Technical Implementation Model accuracy, data handling, and quality of NLP pipeline. (30%)
Interpretability & Output Quality How clear and meaningful are the extracted insights? (25%)
Presentation & Impact Quality of presentation and ease of use. (15%)
02 - Vision For All
Computer Vision Task
Build a computer vision tool to assist visually impaired individuals or enhance
community safety by detecting objects, summarizing scenes, or identifying hazards.
For example: Object detection to read signs in public spaces, scene summarization in
images, identifying hazards in campus environments.
Judging Criteria
Creativity & Use Case Originality of the idea and its societal relevance. (35%)
Technical Execution Model selection, efficiency, and resource optimization. (25%)
Accessibility Impact How well the solution can help the target users. (25%)
Presentation & Impact Quality of presentation and ease of use. (15%)
Page | 2
h ckAIhon
l
03 - Community Decisions By Artificial Intelligence Club
Data Analytics
Participants will use the Urban Air Quality dataset to build an AI-powered tool that help
city planners, agencies, or even individuals make informed decisions to improve air
quality.
Instead of only predicting pollution levels, teams must deploy their insights creatively—
whether through an early warning system, an interactive dashboard for real-time
pollution monitoring, or an app that suggests cleaner commute routes based on air
quality.
Judging Criteria
Insightfulness How valuable and impactful are the discovered patterns. (30%)
Model Accuracy & Approach Model correctness, handling missing data, and clean pipeline. (30%)
Data Storytelling Quality of visualization and clarity of communicated results. (25%)
Presentation & Impact Quality of presentation and ease of use. (15%)
Rules and Deliverables
Choose one of the detailed Challenges above and make sure to register your team
and chosen challenge by 17th April at 11:59 pm.
Teams should consist of 2-3 students and individual submissions are discouraged.
Participants will be required to present their solution in a strictly 7-minute
presentation, explaining their approach, model, key findings from the analysis and
small solution demo.
This applies to all Challenges.
For any further inquiries, kindly contact ai.club@ku.ac.ae or 100063450@ku.ac.ae
Page | 3
