For the second time, we successfully co-organized the international workshop on Interactive Adaptive Learning at this year’s ECML PKDD in Dublin. The goal was to bring together research and application from field. The first workshop took place in Skopje last year and was attended by 100 people. It was a great success! We had 4 hours of tutorial, four presentations, and an invited talk held by Bartosz Krawczyk.
This year, we focused on the discussion of papers and posters. Hence, there was no tutorial but two ignite talks which stated interesting and challenging questions like: How does machine learning influence our environment, our cognitive processes, and our everyday life? Which are the challenges that have to be overcome in order to learn like humans?
The papers presented at the workshop addressed
- Labeling Correctness in Computer Vision Datasets
- Active Learning in Video Scene Classification
- Active Stream Learning for Sentiment Prediction
- Algorithm Recommendation with Active Meta Learning
- Alignment-Based Topic Extraction Using Word Embedding
Also, we were able to introduce our newly developed mobile app Separatio which is available in both, the Play Store, and the App Store. The demo visualized the particular challenges of active learning like selecting the best instances and deciding when to stop learning. Also, 40 out of 60 participants downloaded the app and played at least one game. Some even managed to play the complete first level until the next morning.
Following the paper and poster presentations, the discussion was stimulated by two ignite talks: Georg Krempl on “Influential Machine Learning” and Vincent Lemaire on “Active Learning with transactional data”. After an insightful debate, we concluded that an important step towards interactive adaptive systems has been made, but the repercussions of letting AI systems decide in every aspect of our lives needs to be further investigated.
We are looking forward to the next year’s workshop, which is a great opportunity to participate in the innovation process and meet inspiring persons.