Skip to main content
School of Computer Science and Informatics

Learning and reasoning in artificial intelligence

Professor Thomas Lukasiewicz from the University of Oxford will report on the intersection of learning and reasoning in artificial intelligence (AI).

Location: Room C/2.07, Queen's Buildings, School of Computer Science and Informatics
Date: 25 April 2018 at 14:00
Chair: This seminar will be chaired by Hélène de Ribaupierre.

Abstract

Ontological query answering and reasoning are central problems in many fields, including databases in the form of ontology-based data integration and access; in Big Data management, often along with ontology-based data extraction, integration, and cleaning;  in the Semantic Web, especially for dealing with Linked Data and knowledge graphs; in artificial intelligence (AI) in knowledge representation and reasoning; as well as in many application areas such as medicine, biology, and engineering, where dealing with domain ontologies and knowledge plays an important role.

Seemingly unrelated to ontological query answering and reasoning, a new machine learning technology based especially on neural networks, called deep learning, is recently achieving revolutionary results in speech recognition, visual object recognition, language-related tasks (such as machine translation, language modelling, and sentiment analysis), game playing (such as Google DeepMind’s tackling Atari computer games and the ancient Chinese game of Go), and self-driving vehicles.

However, it has surprisingly turned out that deep-learning-based neural networks can also be used as (offline) compilation of ontological knowledge bases. Experimental results, comparing the novel deep learning system with one of the best logic-based ontology reasoners at present on very large standard benchmarks, show the very promising result that the deep learning system has a very high reasoning accuracy, while being up to two orders of magnitude quicker in (online) query answering.

Biography

Thomas Lukasiewicz is a Professor of Computer Science in the Department of Computer Science at the University of Oxford and a Turing Fellow at the Alan Turing Institute in London, UK.

Prior to this, he held a prestigious Heisenberg Fellowship by the German Research Foundation (DFG), affiliated with the University of Oxford, TU Vienna, Austria, and Sapienza University of Rome, Italy.

Professor Lukasiewicz's research interests are in artificial intelligence (AI) and information systems, particularly:

  • knowledge representation and reasoning
  • uncertainty in AI
  • machine/deep learning
  • personalised search and recommender systems
  • natural language processing and question answering.

He has received the IJCAI-01 Distinguished Paper Award, the AIJ Prominent Paper Award 2013 and the RuleML 2015 Best Paper Award. He is area editor for the journal ACM TOCL, associate editor for the journals JAIR and AIJ, and editor for the journal Semantic Web.