Session #1: June 7, 2022 | 12pm - 1pm ET
This project is a collaboration between Aggregate Intellect, McGill University, and Ryerson University on explainable information retrieval. The goal is to create open source libraries and publish papers on the topics of post-hoc or embedded explainability in Info Retrieval, Search, or related tasks.
Pre-Requisites: There were will be many tasks for various levels of expertise, but strong familiarity with coding in Python, and Machine Learning is expected. Basic familiarity with PyTorch, Git, Deep NLP, GNN, KG-enhanced Search, ElasticSearch and similar techniques and tools is desired.
Leads / Advisors
~ Dr. Amir F. | Advisor
CEO, Aggregate Intellect
~ Dr. Alice Rueda | Advisor
Postdoctoral Fellow at St. Michael's Hospital, Director of ML a Aggregate Intellect
~ Dr. Afshin Amini | Lead
Researcher at The University of British Columbia
~ Percy Chen | Lead
PhD Candidate at McGill University
Project Team
- To be selected after week #3. More info under ‘Working Group Journey’.
❗ Meeting Link and details: https://lu.ma/aisc-xir (add the series to your calendar)
❗ Weekly Meeting Time: 12-1 PM ET on Tuesdays
❗ Slack channel [Communication point]: click here to join
Overview
- Information retrieval and search systems normally use various techniques to generate candidates and then to rank them. Users’ trust of the shortlisting and then ranking process has a significant impact on their willingness to use the system.
- The goal of this project is to explore various post-hoc and embedded methods that can be used to introduce explainability to systems like this. The group will then move on to implement a few potential solutions, and package those as open source libraries.
- Based on the quality of the outcome, we might consider producing some publications.
What will I have to show at the end of this project?
- Contributions to the projects GitHub repo
- Highlights in project recipes, blog posts, social media shoutouts
- [assuming we get some interesting results] authorship of a resultant paper
Prep Steps
- Bookmark the Working Group’s landing page. All key info centralized there.
- Join our Working Group’s Slack channel by:
- By clicking here.
- If having trouble joining, drop ‘ammar@ai.science’ an email. He’ll add!
- [STRONGLY RECOMMENDED BUT OPT] Virtual environment - Google Colab.
- Google Colab is essentially a Google Cloud hosted Juypter notebook.
- Synchronizes everyone’s environments for easier collaborative bug fixing.
- Gives free GPU access. Google Colab - How to use it
Working Group Journey [more info is being added as we go]
Session #1:
- Get to know everyone!
- We will introduce the problem & why it matters.
Session #2:
- Discuss how we’re thinking about the solution.
Solution architecture [Coming up]
Soon coming up
Session #3:
- Ask any clarifying questions.
- End of session → Apply to join the Project Team if it’s your cup of tea.
- Energy & competency are both important, but energy is a little more.
Session #4+ [Tentative length: 2 months]:
- Project Team will work towards the core problem.
- Will be recognized by AISC at the conclusion of the project!
- [For your portfolio / blog / Social Media].
Why join?
Aggregate Intellect hosts one of the most diverse ML communities in the world. Over the course of the working group
- You’ll get an immersion into that community & walk out with some cool new friends.
- Update and enhance your knowledge about information retrieval systems, search, and recommendation systems
- Get your hands dirty with writing code and libraries for a very widely used technique in industry