Four UiB papers accepted to IJCAI-2020
The department of informatics at the University of Bergen will be well represented with four accepted papers at the International Joint Conference on Artificial Intelligence.
The International Joint Conference on Artificial Intelligence (IJCAI) is a highly selective conference. Only 12% of this year's submitted papers were accepted, four of which with co-authors from the department of informatics at UiB.
-It is thus a great achievement that our department has four papers accepted to this conference. This shows that our department has established itself in a leading position in artificial intelligence in Norway, says head of department, Pinar Heggernes.
Two of the accepted papers from UiB are from the algorithms research group, whereas the other two are from the machine learning research group.
- Our algorithms research group has long been in the forefront of international algorithms research; it is a great pleasure to see that they now take a leading role also in artificial intelligence. Our machine learning research group is still very young; such a successful achievement at this early stage is very impressive and highly appreciated. I congratulate all contributors and I thank them for placing our department, our university, and Norway, so visibly on the international artificial intelligence research map, continues Heggernes.
- It is particularly impressive that one of the papers has authors only from our department, two young and fresh colleagues, Ana and Cosimo who started their positions at our department this year, as associate professor and PhD student, respectively.
The following papers with co-authors from our department were accepted (UiB authors in cursive):
|"On the Learnability of Possibilistic Theories"
Cosimo Persia, Ana Ozaki
|"Provenance for the Description Logic ELHr"
Camille Bourgaux, Ana Ozaki, Rafael Penaloza, Livia Predoiu
|"Diversity of Solutions: An Exploration Through the Lens of Fixed-Parameter Tractability Theory"
Julien Baste, Michael R. Fellows, Lars Jaffke, Tomáš Masařík, Mateus de Oliveira Oliveira, Geevarghese Philip, Frances A. Rosamond
Sushmita Gupta, Pallavi Jain, Saket Saurabh
The aforementioned Ana Ozaki started her job as an associate professor at UiB only this March.
- We have seen exciting practical advances in machine learning. However, these must be followed by theoretical development to ensure that results observed empirically are indeed correct and reproducible, says Ozaki.
Her research focus is on the theoretical aspects of machine learning and logic, and she has co-authored two of the four accepted papers.
- In "On the Learnability of Possibilistic Theories", we study a way of representing facts and rules where there is a distinction between when they are taken as ground truth (e.g., there is life on earth) and when they are regarded as beliefs (e.g., there is life in other planets). We propose an approach to learn such kind of representation, which captures inconsistencies and partial knowledge.
- In "Provenance for the Description Logic ELHr", we investigate a strategy for detecting which parts of a collection of facts and rules are relevant for the result of a query posed to this collection. These relevant parts are then represented as a polynomial the indicates the origin (i.e., provenance) of the query result. This can be useful to determine, for instance, the level of trust of query results.
- When recruiting Ana, Heggernes says, we have been confident that she is an exceptionally talented and promising young researcher, and we have high expectations of her. The fact that she fulfills our expectations so quickly and so wonderfully, shows that she is an invaluable asset for our department and for our machine learning research group. I know that Ana and her colleagues in the machine learning group will achieve even greater things in the near future.
For inquiries about AI research at the department of informatics at UiB, our contact info can be found here.