Funded Research Projects

Current:

Unlocking the Potential of AI for English Law

 Project Goals: The proposed research will explore the potential and limitations of using artificial intelligence (AI) in support of legal services.
Principal Investigator: John Armour
Co-Investigators: Abigail Adams, Ewart Keep, Thomas Lukasiewicz, Adam Myles Saunders, Jeremias Prassl, Mari Sako, Rebecca Williams
Partners: Allen & Overy, Law Society, LexSnap, Legal Education Foundation, Slaughter and May, South Square, Thomson Reuters
Type: research grant by the ESRC
Funding Amount: £1,164,743.66
Funding Period: 31/12/2018 to 30/12/2020
More

Interpretable and Explainable Deep Learning for Natural Language Understanding and Commonsense Reasoning

 Project Goals: The main goal of this project is to explore, generalise, and integrate deep learning approaches to structured data extraction and to large-scale logic-based reasoning, towards an interpretable and explainable deep-learning approach to human-like understanding and commonsense reasoning in natural language processing, and to investigate its applications in other disciplines, such as healthcare, law, and finance.
Principal Investigator: Thomas Lukasiewicz
Type: flagship project by the Alan Turing Institute
Funding Period: 2018-2020

Neural Networks with Natural Language Explanations

 Project Goals: We address the critical problems of interpretability and explainability of neural network technologies in AI. We propose a novel and general approach of injecting human-intelligible explanations of the desired outputs during training, as well as requiring them at test time. The objectives of this research are as follows: (i) incorporate explanations during the training of neural networks, (ii) design neural networks that provide explanations in addition to the computed output, and (iii) outsource universal sentence representations to be used off-theshelf in downstream tasks involving natural language.
Principal Investigator: Thomas Lukasiewicz
Type: seed funding project by the Alan Turing Institute
Funding Period: 2018-2019
More

FakeNewsRank: A Ranking for Detecting Fake News on the Web

 Project Goals: Towards a technology for detecting fake news, based on own previous research and developed technologies, the goal of this project is to (i) create a proof-of-concept demonstrator for computing a fake news score, denoted FakeNewsRank, for facts, articles, authors, and websites, which measures their likelihood of being fake, and (ii) to also identify any still existing key gaps or challenges. In addition to this, the project will also produce (iii) datasets of sample blog messages and news articles, containing true and fake facts, along with their authors and web addresses, and (iv) background knowledge graphs underlying these datasets.
Principal Investigator: Thomas Lukasiewicz
Type: impact acceleration award by the EPSRC
Funding Period: 2018-2019
More

RealPDBs: Realistic Data Models and Query Compilation for Large-Scale Probabilistic Databases

 Project Goals: Large-scale probabilistic knowledge bases are becoming increasingly important in academia and industry alike. The state of the art to store and process such data is founded on probabilistic database systems, which are widely and successfully employed. For computational efficiency reasons, however, such systems are typically based on strong, unrealistic limitations, such as the closed-world assumption, the tuple-independence assumption, and the lack of commonsense knowledge. This proposal's main goal is to enhance large-scale probabilistic database systems by more realistic probabilistic data models, while preserving their computational properties. We will develop different semantics for such systems, analyse their computational properties and sources of intractability, design practical scalable query answering algorithms, especially those based on knowledge compilation, along with prototype implementations, and experimentally evaluate these algorithms
Principal Investigator: Thomas Lukasiewicz
Co-Investigators: Georg Gottlob, Dan Olteanu
Partners: Wrapidity Limited
Type: standard grant by the EPSRC
Funding Amount: £781,286
Funding Period: 01/12/2017 to 31/05/2021
More

VADA: Value Added Data Systems - Principles and Architecture

 Project Goals: The VADA research programme will define principles and solutions for Value Added Data Systems, which support users in discovering, extracting, integrating, accessing, and interpreting the data of relevance to their questions.
Principal Investigator: Georg Gottlob
Co-Investigators: Oscar Peter Buneman, Wenfei Fan, Alvaro Adolfo Fernandes, Tim Furche, Paolo Guagliardo, John Keane, Leonid Libkin, Thomas Lukasiewicz, Sebastian Maneth, Dan Olteanu, Giorgio Orsi, Norman William Paton, Andreas Pieris
Partners: Neo4j, Facebook, The Christie Hospital Charitable Appeals, UK, LambdaTek, AllianceBernstein plc., FutureEverything CIC, UK, Horus Security Consultancy Ltd, Cheltenham, Huawei Technologies Co Limited, China, Microsoft Research, US, Neo Technology, UK, PricePanda Group, Logicblox
Type: programme grant by the EPSRC
Funding Amount: £4,557,635
Funding Period: 01/04/2015 to 31/03/2020
More

DBOnto: Bridging Databases and Ontologies

 Project Goals: Semantics-aware data management involves a synthesis of ontological reasoning and database management principles. It employs rich schemas (AKA ontologies) that allow to deal with incomplete and semi-structured information from heterogeneous sources, and to answer queries in a way that reflects both knowledge and data, i.e., to deliver understanding from information. The goal of the grant is to develop next-generation semantics-aware data management systems.
Principal Investigator: Ian Horrocks
Co-Investigators: Michael Benedikt, Bernardo Cuenca Grau, Georg Gottlob, Thomas Lukasiewicz, Boris Motik, Dan Olteanu
Type: platform grant by the EPSRC
Funding Amount: £1,263,746
Funding Period: 29/01/2014 to 28/01/2019
More

Previous:

Preference Mining for Personalised Search on the Social Semantic Web

 Project Goals: The main goal of this proposal is to explore and develop suitable techniques for mining preferences for personalised search on the Social Semantic Web. We especially want to extract preferences from textual documents on the Web (such as blog posts, recommendations, and reviews) via deep-learning-based techniques for natural language processing, such as deep-learning-based sentiment analysis. We also want to develop a small preference-mining demonstrator.
Principal Investigator: Thomas Lukasiewicz
Type: EPSRC Doctoral Prize for Oana Tifrea-Marciuska
Funding Period: 2017-2018
More

Intelligent Question Answering

 Project Goals: The main goal of this project is to develop a proof-of-concept prototype implementation for a small natural language question answering system (for pieces of text in medicine), which combines recent deep learning technologies for natural language processing with knowledge representation and reasoning technologies.
Principal Investigator: Thomas Lukasiewicz
Type: seed funding project by the Alan Turing Institute
Funding Period: 2016-2017

Probabilistic Ontologies for Real-World Hypotheses

 Project Goals: Integration of predictions with ontologies, to find representations that are suitable for representing real-world hypotheses, algorithms for inference on these representations, and ways to learn them.
Type: Leverhulme Visiting Professorship for David Poole
Funding Period: 2014-2015

Personalized Search on the Social Web

 Project Goals: Personalized search aims at adapting search results based on the tastes, interests, and needs of the user. In the proposed research, we want to explore the idea of integrating Web search and social networks, both for an improved personalized Web search and for personalized search on the Social Web. In particular, towards building user profiles for personalization, we want to explore forma- lisms and algorithms for mining, representing, and managing user preferences, and for using them in personalized search on the Social Web. We also want to consider the modeling of relevant background knowledge via ontologies, which then allows for personalized semantic search on the Social Web.
Principal Investigator: Thomas Lukasiewicz
Type: Google European Doctoral Fellowship for Oana Tifrea-Marciuska
Funding Period: 2013-2016
More

PRODIMA: Probabilistic Data and Information Integration with Provenance Management

 Project Goals: The project aims at defining a notion of uncertain provenance for probabilistic information integration frameworks in the Semantic Web, at investigating its semantic properties, and at developing scalable reasoning algorithms for collecting, storing, and querying provenance information to facilitate the computation of probabilities and to enable provenance-based mapping debugging.
Type: Marie-Curie Intra-European Fellowship for Livia Predoiu
Funding Period: 2013-2015
More

PrOQAW: Probabilistic Ontological Query Answering on the Web

 Project Goals: Towards next-generation technologies for semantic search and query answering on the Web, the project’s central goal is to develop a family of probabilistic data models, along with scalable query answering algorithms, for knowledge bases that are extracted from the Web relative to an underlying ontology. We want to realize this by integrating ontology languages, database technologies, and formalisms for managing probabilistic uncertainty. The objectives include developing probabilistic data models, developing algorithms for ranking and query answering, identifying useful scalable fragments, and practically evaluating our results.
Principal Investigator: Thomas Lukasiewicz
Type: standard grant by the EPSRC
Funding Period: 2012-2015
More

Probabilistic Semantic Query Answering on the Web

 Project Goals: Towards semantic search and query answering on the Web, the goal of this project is to develop a family of probabilistic data models for knowledge bases extracted from the Web, along with scalable query answering algorithms.
Principal Investigator: Thomas Lukasiewicz
Type: Google Research Award
Funding Period: 2011-2012
More

ExODA: Integrating Description Logics and Database Technologies for Expressive Ontology-Based Data Access

 Project Goals: Towards the next generation of ontology-based ISs, development of formalisms and methods for a synthesis and an extension of ontology and database systems and techniques, with data handling capabilities similar to current RDBMSs, but schemas that are rich, flexible, and tightly integrated with the data.
Type: standard grant by the EPSRC
Funding Period: 2011-2014
More

From Data Extraction to Ontology-Based Semantic Search on the Web

 Project Goals: Development of formalisms and methods for automatic extraction of data from the Web, such as the development of a general model of knowledge representation and reasoning for integrating different sources on the Web, directed towards possible key technologies for the future Web, such as semantic searching.
Type: Yahoo! Research Fellowship
Funding Period: 2010-2013

Databases, Web, and AI

 Project Goals: Development of formalisms and methods for the Web, which combine database techniques for handling very large data with intelligent techniques from AI.
Principal Investigator: Thomas Lukasiewicz
Type: Heisenberg Fellowship by the German Research Foundation (DFG)
Funding Period: 2007-2009

Game-Theoretic Agent Programming

 Project Goals: Development of high-level formalisms and techniques for programming systems of multiple agents in uncertain and partially observable environments.
Principal Investigator: Thomas Lukasiewicz
Type: Research project funded by the Austrian Science Fund (FWF)
Funding Period: 2005-2008
More

Knowledge Representation and Reasoning for Intelligent Agents and the Semantic Web

 Project Goals: Development of KR formalisms and techniques (i) for high-level probabilistic agent programming and reasoning, (ii) for planning under uncertainty in AI, (iii) for the Rules, Logic, and Proof Layers of the Semantic Web.
Principal Investigator: Thomas Lukasiewicz
Type: Heisenberg Fellowship by the German Research Foundation (DFG)
Funding Period: 2004-2007

Expressive Probabilistic Description Logics and their Integration into Probabilistic Databases

 Project Goals: Development of rich probabilistic description logics, and exploration of their computational properties. Integration of these probabilistic description logics into probabilistic databases for an increased expressivity.
Principal Investigator: Thomas Lukasiewicz
Type: Marie Curie Individual Fellowship by the European Union
Funding Period: 2001-2004

Probabilistic Logic Programming and Many-Valued Logic Programming with Probabilistic Semantics

 Project Goals: Development of approaches to probabilistic logic programming and many-valued logic programming with probabilistic semantics, which allow for disjunctions in rule heads and negations in rule bodies. Analysis of the computational complexity of probabilistic logic programming and many-valued logic programming with probabilistic semantics, and elaboration of efficient algorithms for these tasks.
Principal Investigator: Thomas Lukasiewicz
Type: Habilitation Fellowship by the German Research Foundation (DFG)
Funding Period: 1999-2001


© Copyright 2018. All Rights Reserved.