Are you tired of the endless search for the perfect gift online? It’s like navigating a maze of products, reviews, and prices, only to feel overwhelmed by too many choices.
Introducing the Amazon Multi-Task Online Shopping Challenge, where we’re revolutionizing online shopping using Large Language Models (LLMs). Traditional methods miss the mark in understanding the nuance of shopping terminology, consumer behavior, and the wide array of products and languages, leaving users drowning in information.
Our ShopBench benchmark mirrors real-world shopping complexities, aiming to make online shopping as intuitive as having a knowledgeable assistant by your side. Participate to develop LLMs that can simplify shopping, making it a more intuitive and satisfying experience, much like a knowledgeable shopping assistant would in real life.
Multi-Task Online Shopping Challenge for LLMs
With 57 tasks and over 20,000 questions based on real Amazon data, this challenge pushes LLMs to excel in understanding shopping concepts, customer behavior, and multilingual support. Whether you’re a seasoned developer or a student, from the industry or academia, this challenge offers a platform to craft innovative LLM solutions that reshape online shopping experiences and valuable insights that benefit the whole community.
ShopBench, a comprehensive benchmark that mimics these real-world online shopping complexities, focuses on four main key shopping skills (which will serve as Tracks 1-4):
- shopping concept understanding
- shopping knowledge reasoning
- user behavior alignment
- multi-lingual abilities
Additionally, Track 5: All-around, promotes comprehensive solutions that address all tasks in Tracks 1-4 with a single, unified approach, offering larger rewards for these versatile solutions.
This challenge aims to give participants practical experience in crafting advanced LLM solutions for real issues, benefiting both the online service industry with robust, ready-to-implement LLM solutions and the wider machine learning community with valuable insights and training guidance.
Exciting Prizes
The challenge offers a total prize pool of $41,500, divided into three categories:
- Winner Prizes: Cash awards for the top three positions in each track.
- AWS Credits: Awarded to teams ranking immediately after the top three in each track.
- Student Awards: Special awards for the best student teams to support the development of resource-efficient LLM solutions due to the high computational costs and engineering efforts involved.
Prizes for Tracks 1-4:
-
First place: $2,000 -
Second place: $1,000 -
Third place: $500 - 4th-7th places receive AWS Credit of $500
-
Student Award: $750
Prizes for Track 5 (All-around):
-
First place: $7,000
-
Second place: $3,500
-
Third place: $1,500
- 4th-8th places receive AWS Credit of $500
-
Student Award: $2,000
Winners have the opportunity to present their work at the KDD Cup workshop 2024, held at ACM SIGKDD 2024 (August 2024, Barcelona, Spain).
Challenge Timeline
- Phase 1 Start Date: 21th March, 2024 23:55 UTC
- Entry Freeze Deadline and Phase 1 End Date: 10th May, 2024 23:55 UTC
- Phase 2 Start Date: 15th May, 2024 23:55 UTC
- End Date: 10th July, 2024 23:55 UTC
- Winner Notification: 15th July, 2024
- Winner Announcement: 26th August, 2024 (At KDD 2024)
Signup now to begin this journey and dive into the challenge details. Join a community of innovative thinkers, share ideas, and engage in this exciting challenge.
Challenges are more fun with teams. Find your teammate.
Have feedback or query? Share it with us.
All the best,
Team AIcrowd