Intelligent Systems

The Intelligent Systems Group’s research goal is to develop more intelligent, autonomous, accountable, and trustworthy systems.

Despite old expectations that computational systems behave intelligently, only more recently have we begun to witness the true potential of Artificial Intelligence, partly because of novel AI techniques, but mostly by relying on scaling data sizes, model capacity, and computational power.
However, as AI becomes pervasive, old problems resurface and new challenges arise.
As we exhaust human-generated data, progress towards robust, trustworthy AI will require the exploitation of domain knowledge in Machine Learning, the extraction of symbolic actionable knowledge from ML models, and ultimately merging data-based machine learning with knowledge-based reasoning.

Building upon the long-standing tradition of being the first Portuguese research group working in AI, the IS group results from the realignment of the previous Knowledge-Based Systems group through the addition of experts in ML to face the current challenges of AI and work towards this long-term vision.

The IS group focuses its research on a) the development of models whose components can mechanistically encode domain knowledge in training and inference as inductive biases, such as physical constraints or domain-elicited knowledge such as rules, so that the model has just enough capacity to learn each task; b) the development of sound ways to help humans interpret current deep models, new models that humans can interpret at different scales, which can be safely trusted, as well as mechanisms to generate knowledge from these models and effectively use it in combination with existing domain knowledge; and c) the development of novel knowledge representation languages and reasoning procedures adequate to represent and manipulate knowledge in a way that is formal, efficient, and human understandable, while being exploitable in this new age of AI.

The IS group collaborates with other international and national researchers and practitioners, with a strong focus on technology transfer in the areas of Bioinformatics, Aerospace, Natural Resources Management, and Robotics.

The IS group is a leading national research group in AI with a strong international reputation in Knowledge Representation and Reasoning (KRR) and Machine Learning (ML). The group, which includes an ECCAI Fellow, regularly publishes at the top journals such as the AI Journal, ACM TOCL, JAIR, TNNLS, and TPLP, and top conferences such as IJCAI, AAAI, KR, AAMAS, ICML, and ISWC, to whose PCs it regularly belongs.