http://www.cs.ox.ac.uk/projects/ All en-gb 360 Mon, 17 Jun 2013 16:17:52 GMT Mon, 17 Jun 2013 16:17:51 GMT All http://blogs.law.harvard.edu/tech/rss/ http://www.cs.ox.ac.uk/projects/safesea/ <p> <div>SAFeSEA (Supportive Automated Feedback for Short Essay Answers) is a new EPSRC-funded research project&nbsp;between the Institute of Educational Technology at the UK&rsquo;s Open University and the Computer Science Department at Oxford University.</div> <div>&nbsp;</div> <div>The aim of this project is to provide an effective automated interactive feedback system that yields an acceptable level of support for university students writing essays in a distance or e-learning context.</div> <div>&nbsp;</div> <div>The project will explore the development of natural language processing tools for essay summarising, the definition of effective models of feedback and the design of web-based wrappers to orchestrate tools and evaluate different supportive approaches.</div> </p> Department of Computer Science, University of Oxford Projects All oucl-All-safesea Sat, 30 Mar 2013 09:29:07 GMT http://www.cs.ox.ac.uk/projects/PRODIMA/ <p>The PRODIMA project investigates provenance-based probabilistic information integration in the Semantic Web. A notion of uncertain provenance will be defined and investigated for usage in probabilistic information integration frameworks suitable in the Semantic Web context. Properties of this notion of uncertain provenance will be formally and logically investigated. Furthermore, it is aimed to develop scalable reasoning algorithms for collecting, storing, and querying provenance information, in order to accelerate the computation of probabilities, but also in order to facilitate provenance-based mapping debugging. By this means, it will be possible to combine reasoning, mapping debugging, and provenance management through scalable algorithms. At the end of the project, a small demo system will showcase the benefits of provenance-based probabilistic information integration in the Semantic Web context. &nbsp;</p> Department of Computer Science, University of Oxford Projects All oucl-All-PRODIMA Fri, 22 Mar 2013 13:16:58 GMT http://www.cs.ox.ac.uk/projects/CITD/ <p>It is widely recognised that the threat to enterprises from insider activities is increasing and that significant costs are being incurred. The multi-faceted dimensions of insider threat and compromising actions have resulted in a diverse experience and understanding of what insider threats are and how to detect or prevent them. The purpose of this research is to investigate the potential for near real-time detection of insider threat activities within a large enterprise environment using monitoring tools centred around the information infrastructure. As inside threat activities are not confined solely to cyber-based threats, the research will explore the potential for harnessing a variety of threat indicators buried in a different enterprise operations connected or interfacing with the information infrastructure, while enabling human analysts to make informed decisions efficiently and effectively.</p> <p><strong>PROJECT OVERVIEW</strong></p> <p class="Default">Our research incorporates both theoretical and applied research aimed at delivering a significantly enhanced capability in insider threat detection, as well as education and dissemination materials and strategies designed to maximise uptake of the insight generated by the research. Our approach is to combine cyber security, psychology, criminology, visual analytics, enterprise operations management and executive education expertise to:</p> <ul> <li> <div class="Default">Develop a model for insider threat which is flexible enough to underpin detection systems based on both detecting deviations from normal behaviour, and the identification of specific events of interest which might indicate the presence of an attack involving an insider. The model will support the distinguishing of attack events relating to activities in the physical space and cyber space, based on data sources accessible via the information infrastructure.</div> </li> <li> <div class="Default">Understand the potential for psychological indicators of an insider becoming a threat, including how we might detect such indicators based on cyber behaviours.</div> </li> <li> <div class="Default">Identify the most effective pattern extraction algorithms for facilitating correlation and detection across heterogeneous operational contexts.</div> </li> <li> <div class="Default">Understand the enterprise culture and common practices that such novel detection systems would need to work within, and design processes appropriate to enabling operation.</div> </li> <li> <div class="Default">Provide a visual analytical interface to assist human analysts in more complex reasoning and decision-making processes by enabling them to fuse their knowledge and experience with the information and threat indicators discovered by the system, hence empowering the analysts to play an active role within the detection system in addition to being consumers of its outputs.</div> </li> <li> <div class="Default">Develop an understanding of both the various organisational roles that will be impacted by such an insider threat detection system and have responsibilities towards successful outcomes, and the various awareness raising and educational methods which are likely to have the greatest impact in enabling stakeholders to benefit from the research and to learn from the knowledge developed.</div> </li> </ul> <p>We will be working closely with Financial Fraud Action UK, SOCA, CISCO, CIFAS - the UK's Fraud Prevention Service - and others.</p> Department of Computer Science, University of Oxford Projects All oucl-All-CITD Mon, 26 Nov 2012 17:16:49 GMT http://www.cs.ox.ac.uk/projects/Optique/ <p>Scalable end-user access to Big Data is essential for the effective support of critical decision making in large companies. The Optique project aims to develop new techniques and infrastructure that will bring about a paradigm shift for data access by:</p> <ul> <li>using Ontology Based Data Access (OBDA) to provide a semantic end-to-end connection between users and data sources;</li> <li>enabling users to rapidly formulate intuitive queries using familiar vocabularies and conceptualisations;</li> <li>seamlessly integrating data spread across multiple distributed data sources, including streaming sources;</li> <li>exploiting massive parallelism for scalability far beyond traditional RDBMSs;</li> </ul> <p>and thus reducing the turnaround time for information requests to minutes rather than days.</p> <p>These objectives will be achieved by bringing together leading researchers and developers from diverse communities &mdash; including Knowledge Representation, Databases, and the Semantic Web &mdash; to devise new techniques and to implement them in an extensible platform that will provide a complete and generic solution to the data access challenges posed by Big Data.</p> <p>The platform will: (i) Use an ontology and declarative mappings to capture user conceptualisations and to transform user queries into complete, correct and highly optimised queries over the data sources; (ii) Integrate distributed heterogeneous sources, including streams; (iii) Exploit massively parallel technologies and holistic optimisations to maximise performance; (iv) Include tools to support query formulation and ontology and mapping management; and (v) Use semi-automatic bootstrapping of ontologies and mappings and query driven ontology construction to minimise installation overhead.</p> <p>Development of the platform will be informed by and continuously evaluated against the requirements of complex real-world challenges, with two large European companies providing the project with comprehensive use cases, and access to user groups and TB scale data sets.</p> Department of Computer Science, University of Oxford Projects All oucl-All-Optique Tue, 10 Jul 2012 15:07:28 GMT http://www.cs.ox.ac.uk/projects/Score/ <p>Decisions in industry, government and health care increasingly depend on improved access to and processing of digital information. This has led to a pressing demand for more powerful and flexible information systems. New generation information systems will need to efficiently process large data sets, exploit machine-readable domain knowledge, and answer queries by taking into account both knowledge and data.</p> <p>Ontology-based information systems (OISs) constitute a rapidly maturing technology with the potential to meet these requirements. An ontology provides a vocabulary of terms that are familiar to the user, together with axioms describing the meaning of those terms. OISs can exploit the rich domain knowledge in an ontology to provide a unified view of the data and enrich query answers with implicit information using an automated reasoner.</p> <p>Several standards for ontology and query languages have been developed, including RDF, OWL, OWL 2, and SPARQL. OWL and its revision OWL 2 provide a powerful and flexible ontology modelling language that can capture features such as class hierarchies, incomplete information, negative information, and so on. OWL ontologies are being used in an increasing range of applications, and are becoming a core technology for accessing, gathering, and sharing knowledge and data.</p> <p>Applications involving large amounts of data, however, still pose serious challenges to the applicability of OISs. Problems in the applicability of OISs typically originate from conflicting application requirements.</p> <ul> <li> Modelling complex application domains requires rich ontology languages. </li> <li> Fine-grained access to information requires powerful query languages. </li> <li> Answering queries over large data sets requires scalable reasoners. </li> <li> Critical decisions that depend on access to information require query answers that are either complete, or where the incompleteness is well-understood. </li> </ul> <p>Due to high worst-case complexity of the relevant reasoning problems, scalability is usually in conflict with the use of powerful ontology and query languages, and many applications give up completeness to achieve the desired scalability. As a result, existing OISs fail to meet one or more of these requirements: they support only weak ontology or query languages, they do not scale to the required volumes of data, or they do not provide guarantees as to the completeness of query answers.</p> <p>Our goal in this project is to lay the foundations for a new generation of OISs that meet all the aforerementioned requirements, thus providing the ideal combination of expressive power, scalability and completeness. To accomplish such an ambitious goal, we observe that the limitations imposed by the trade-offs between expressivity, scalability and completeness apply at the language level: that is, they involve worst-case complexity bounds for every ontology, query, and data set expressed in given ontology, query and data modeling languages.</p> <p>The class of ontologies, queries and data sets that are relevant to a particular application is, however, much more restricted. For example, although application data is often unknown or frequently changing, the ontology itself is fixed at design time, or changes infrequently. As a result, a reasoner known to be incomplete in general for given query and ontology languages might yield the same results as a complete reasoner for the application at hand. Identifying such cases is challenging, but it would have tremendous added value: applications could exploit scalable incomplete reasoners while still enjoying completeness guarantees, thus achieving 'the best of both worlds'.</p> <p>We believe that our main goal can be accomplished by designing OISs that are optimised for the ontologies, queries and data sets relevant to the application at hand. Such OISs would maximise scalability while ensuring completeness of query answers, even for rich ontologies, large-scale data sets, and complex user queries.</p> Department of Computer Science, University of Oxford Projects All oucl-All-Score Mon, 09 Jul 2012 21:44:46 GMT http://www.cs.ox.ac.uk/projects/FDB/ <P>Factorised databases are relational databases that use compact factorised representations at the physical layer to reduce data redundancy and boost query performance. They are based on algebraic factorisation using distributivity of relational product over relational union and commutativity of product and union. <BR> The goal of the project is to understand the benefit that factorised representations can bring to query evaluation in relational databases, and in particular <STRONG>build a relational data management system that uses factorisations at the physical layer and relations at the logical layer</STRONG>.</P> <P>Our initial investigation quantifies the gap between sizes of relations and of their factorised representations. In an ICDT 2012 paper we give two characterisations of conjunctive queries based on factorisations of their results whose nesting structure is defined by so-called factorisation trees. The first characterisation concerns sizes of factorised representations. For any query, we derive a size bound that is asymptotically tight within our class of factorisations. We also characterise the queries by tight bounds on the readability of the provenance of result tuples and define syntactically the class of queries with bounded readability. Future work is to investigate the space of possible compact representation systems for relational data and quantify the trade-off between their succinctness and tractability of query evaluation.</P> <P>We are currently building <STRONG>FDB</STRONG>, an in-memory query engine on factorised databases. Key components of FDB are novel algorithms for query optimisation and evaluation that exploit the succinctness brought by data factorisation. Preliminary experiments show that for data sets with many-to-many relationships FDB can outperform relational engines by orders of magnitude.</P> <H2>News</H2> <UL> <LI><STRONG>September 2012</STRONG>: Tomáš Kočiský was awarded the Gloucester Research Project Prize for best 4th year Maths&CS project 2012. His project investigates the evaluation problem of queries with ORDER-BY and GROUP-BY clauses on factorised databases. </LI> <LI><STRONG>June 2012</STRONG>: Jakub Závodný will do an internship over the summer of 2012 at Google Zurich.</LI> </UL> <H2>System prototype</H2> <P>We are currently working on a second, improved version of FDB to be released later in the year. If you want to try out the current version, please contact a member of our team.</P> <H2>Talks</H2> <UL> <LI><EM>Factorised Relational Databases: Results So Far and Open Questions</EM><BR> A series of 3 invited lectures (each 45 minutes) given by Jakub at Olomouc University (September 2013, Czech Republic) </LI> </UL> <UL> <LI><EM> Factorised Relational Databases</EM><BR> 1 hour talk given by Jakub in Constraints Seminar (Oxford, March 2012) and by Dan in: Systems Seminar (EPFL, Feb 2012, Lausanne) and Birkbeck Computer Science Departmental Seminar (March 2012, London). </LI> </UL> <UL> <LI><EM> Factorised Representations of Query Results</EM><BR> 1 hour talk given by Jakub in Information Systems Seminar (Oxford, November 2011) and by Dan in: Database Seminar (University of Edinburgh, March 2011, Edinburgh) and Dagstuhl Seminar on "Foundations of Distributed Data Management" (Oct 2011, Dagstuhl). </LI> </UL> <H2>Theses</H2> <UL> <LI>Laura Drahici: <EM>Updates in Factorised Databases.</EM><BR> MSc in CS, Oxford 2013 (ongoing).</LI> </UL> <UL> <LI>Szymon Wyleżoł: <EM>Cost-based Query Optimisation for Factorised Databases.</EM><BR> MSc in Maths and CS, Oxford 2012.</LI> </UL> <UL> <LI>Tomáš Kočiský: <EM>Queries with Aggregates and Order-By on Factorised Databases.</EM>[<A href="http://www.cs.ox.ac.uk/dan.olteanu/theses/Tomas.Kocisky.pdf">pdf</A>]<BR> MSc in Maths and CS, Oxford 2012. </LI> </UL> <UL> <LI> Nurzhan Bakibayev: <EM>A Query Engine for Factorised Databases</EM><BR> MSc in CS, Oxford 2011. </LI> </UL> <H2>Publications</H2> <UL> <LI><STRONG>Aggregation and Ordering in Factorised Databases</STRONG>. <BR>Nurzhan Bakibayev, Tomáš Kočiský , Dan Olteanu, and Jakub Závodný. <BR>March 2013. <BR> Complementary information available in <A href="http://www.cs.ox.ac.uk/dan.olteanu/theses/Tomas.Kocisky.pdf">Tomas' thesis</A></LI> </UL> <UL> <LI><STRONG>Demonstration of the FDB Query Engine for Factorised Databases</STRONG>. [<A href="http://www.cs.ox.ac.uk/dan.olteanu/papers/boz-vldb12b">pdf</A>; <A href="http://www.cs.ox.ac.uk/dan.olteanu/papers/vldb12-poster.pdf">poster</A>] <BR>Nurzhan Bakibayev and Dan Olteanu and Jakub Závodný. <BR>System demonstration. In Very Large Data Bases (PVLDB), 5(12), 2012. Istanbul, 2012. </LI> </UL> <UL> <LI><STRONG>FDB: A Query Engine for Factorised Relational Databases</STRONG>. [<A href="http://www.cs.ox.ac.uk/dan.olteanu/papers/boz-vldb12">pdf</A>] <BR>Nurzhan Bakibayev and Dan Olteanu and Jakub Závodný. <BR>In Very Large Data Bases (PVLDB), 5(12), 2012. Istanbul, 2012. <BR>Prior version in <A href="http://arxiv.org/abs/1203.2672">arXiv technical report abs/1203.2672.</A></LI> </UL> <UL> <LI><STRONG>Factorised Representations of Query Results: Size Bounds and Readability</STRONG>. [<A href="http://www.cs.ox.ac.uk/dan.olteanu/papers/oz-icdt12.pdf">pdf</A>, <A href="http://www.cs.ox.ac.uk/dan.olteanu/tutorials/fdb.pdf">slides</A>] <BR>Dan Olteanu and Jakub Závodný. <BR>In Int Conf on Database Theory (ICDT), Berlin, 2012. <BR> Prior version (strict subset of results) in <A href="http://arxiv.org/abs/1104.0867"> arXiv technical report 1104.0867</A>, April 2011. </LI> </UL> <UL> <LI><STRONG>On Factorisation of Provenance Polynomials</STRONG>. [<A href="http://www.cs.ox.ac.uk/dan.olteanu/papers/oz-tap11.pdf">pdf</A>, <A href="http://www.cs.ox.ac.uk/dan.olteanu/papers/factorisation-poster2011.pdf">poster</A>] <BR>Dan Olteanu and Jakub Závodný. <BR>In 3rd USENIX Workshop on the Theory and Practice of Provenance (TaPP), June 2011, Heraklion, Crete. </LI> </UL> <H2>Current Team</H2> <UL> <LI>Laura Draghici</LI> <LI>Dan Olteanu</LI> <LI>Jakub Závodný</LI> </UL> <P>Alumni: Nurzhan Bakibayev (MSc in CS, 2011); Szymon Wyleżoł (4th year Maths&CS, 2012); Tomáš Kočiský (4th year Maths&CS, 2011-2012);</P> <H2>Acknowledgments</H2> <P>Jakub Závodný is supported by an EPSRC DTA Grant EP/P505216/1. From June 2011 to April 2012, Jakub was partially supported by the Future and Emerging Technologies (FET) programme within the Seventh Framework Programme for Research of the European Commission, under the FETOpen grant agreement FOX, number FP7-ICT-233599.</P> Department of Computer Science, University of Oxford Projects All oucl-All-FDB Thu, 21 Jun 2012 15:52:51 GMT http://www.cs.ox.ac.uk/projects/RACE/ <p>RACE is a five-year ERC Advanced Grant, aimed at developing the techniques required to understand, model, and analyse the equilibrium properties of game-like distributed systems. Building on two decades of highly successful research by the principal investigator, RACE will carry out research at the intersection of game theory, logic, complexity, and verification.</p> Department of Computer Science, University of Oxford Projects All oucl-All-RACE Wed, 13 Jun 2012 17:02:43 GMT http://www.cs.ox.ac.uk/projects/PrOQAW/ <p><span>The next revolution in Web search as one of the key technologies of the Web has just started with the incorporation of ideas from the Semantic Web, aiming at transforming current Web search into some form of semantic search and query answering on the Web, by adding meaning to Web contents and queries in the form of an underlying ontology. This also allows for more complex queries, and for evaluating queries by combining knowledge that is distributed over many Web pages, i.e., by reasoning over the Web. </span><br /><br /><span>Realizing such semantic search and query answering on the Web by adding ontological meaning to the current Web conceptually means annotating Web pages and their contents relative to that ontology, i.e., relating Web pages and their contents to and thus also via that ontology. From a practical perspective, one of the most promising ways of realizing this is to perform data extraction from the current Web relative to the underlying ontology, store the extracted data in a knowledge base, and realize semantic search and query answering on this knowledge base. There are recently many strong research activities in this direction.</span><br /><br /><span>A major unsolved problem in the above context is the principled handling of uncertainty: In addition to natural uncertainty as an inherent part of Web data, one also has to deal with uncertainty resulting from automatically processing Web data. The former also includes uncertainty due to incompleteness and inconsistency in the case of missing and over-specified information, respectively. The latter includes uncertainty due to, e.g., the automatic annotation of Web pages and their contents, the automatic extraction of knowledge from the Web, matching between different related ontologies, and the integration of distributed Web data sources.</span><br /><br /><span>The central goal of the proposed research is to develop a family of probabilistic data models for knowledge bases extracted from the Web relative to an underlying ontology, along with scalable query answering algorithms, which may serve as the backbone for next-generation technologies for semantic search and query answering on the Web. We believe that such probabilistic data models and query answering algorithms can be developed by integrating ontology languages, database technologies, and formalisms for managing probabilistic uncertainty in the context of the Web. The objectives include developing probabilistic data models, developing algorithms for ranking and query answering, identifying useful scalable fragments, and practically evaluating our results.<br />&nbsp; </span></p> Department of Computer Science, University of Oxford Projects All oucl-All-PrOQAW Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/EPSRCUAVAutonomy/ <p><span id="lblAbstract" style="font-weight: normal;">In March 2011, Japan suffered from its biggest earthquake and devastating tsunami. Severe damage were inflicted on its Fukushima nuclear plants and more than 100,000 people had to be evacuated after the radiation levels became unsafe. Workers were not able to operate on site, preventing them from securing safety at the atomic power plant and averting a major radiation leak. One month after the disaster, in order to assess the severity of the damage to the nuclear plant from above, a small aerial vehicle equipped with cameras was sent to take pictures and videos of the affected areas. The video footage obtained brought valuable information to the rescue teams that could not have been acquired otherwise. But the use of aerial vehicles still remains limited by the fact that they require a remote operator at transmission range to control them. It is also necessary to have an operator to control the camera and interpret the data.<br /><br />In order to work autonomously, these systems need to be highly intelligent and rational so that they can become reliable: they must have high levels of knowledge to accomplish their AI-complex missions which occur in any other information environment. This implies that they should adapt to any unexpected situations such as recent changes not reflected in prior information on the environment and possible loss of GPS due to obstructing buildings or indoor exploration; reliable operation under such conditions would, for instance, enable them to return safely to their base station. In a multi-UAV setting, they should additionally be able to communicate with each other to simplify their goals, to learn from each other's information, and to update and share their knowledge. Given that any mission is unique in terms of deployment areas, tasks and goals to be achieved, etc., and can be critical in the sense that human lives may be involved, the implementation must be verified to be correct with respect to a formal specification. A famous example of an implementation error and a failure to comply with the specification is the self-destruction of Ariane 5 in 1996 immediately after take-off, caused by a numeric overflow due to an implementation that was not suitable for all possible situations. In 1996, the Lockheed Martin/Boeing Darkstar long-endurance UAV crashed following what the Pentagon called a "mishap [..] directly traceable to deficiencies in the modelling and simulation of the flight vehicle".<br /><br />To achieve the reliability required, we will need to develop a formalism that represents the sets of actions each Unmanned Aerial Vehicle (UAV) can perform while allowing capture of the kinetic constraints of the UAVs. We will then verify that the behaviours of each UAV modelled using this formalism lead to the individual or overall goal of the mission they are to achieve. These need to be extended from individual behaviours to a cooperative level amongst the multiple UAVs. Next, we plan to link the low-level code to high-level abstraction and verify it via advanced model-checking techniques. Finally, logical tools will be used to exhaustively reason about learning as a result of information flow among UAVs and their environment.</span></p> Department of Computer Science, University of Oxford Projects All oucl-All-EPSRC UAV Autonomy Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/SIPCS/ <p>This project aims to fund a network to support an emerging research community working at the interface of Computer Science, Physics and Mathematics. The common theme is the use of high-level mathematical structures such as Category Theory. These structures have already proved important in various specific areas within these disciplines. It has recently been becoming clear that they can be used in a unified and very powerful way, and that the same structures occur across these disciplines. This has led to a lot of exciting research, which has been bringing established researchers together, and attracting students and young researchers into the area.<br /><br />The aim of the network is to foster and support these activities, and to nurture the development of an emerging research community, centred in the UK, which is at the forefront of these exciting developments.<br /><br />We will do this by supporting regular meetings, providing travel money for young researchers to attend conferences and make research visits, and managing a strong web presence which can act as a hub for the community. This will help to make the UK attractive to the best researchers in the world on the topic of the network, and will enable it to take a leading role in future developments.</p> Department of Computer Science, University of Oxford Projects All oucl-All-SIPCS Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/UnderTracker/ <p>Wildlife tracking using wireless sensor networks has garnered a great deal of attention and research, since the seminal ZebraNet project monitored zebras with mobile nodes in 2002. However, research to date has concentrated on monitoring animals when they are above ground. It is currently impossible to automatically monitor animals whilst they are underground. The main reason for this is that radio waves are severely attenuated by layers of soil, to the point of being unusable. There is a strong need for a system that can localize burrowing animals when they are within their dens or tunnels, in order to better understand their behaviour and habits. A prime example of this is the European badger - badgers are a protected species in the UK, yet are subject to widespread culling due to their possible link to bovine TB. By monitoring internal sett conditions and animal interactions underground, a better understanding of infection could potentially be obtained. <br /><br />To tackle these issues, I propose the use of low frequency magnetic fields (i.e. the principle of magneto-induction, MI), which are able to penetrate soil without attenuation, to provide ultra-low power three dimensional localization of wild animals within their burrows. Data from tracking collars will be forwarded by conventional high frequency radio links when the animal is above ground, meaning that the animal does not need to be recaptured to obtain the stored information. By mapping animal movements over time, the subterranean tunnel architecture itself will be determined, something which can currently only be obtained, destructively, through excavation. Sensors within the tunnel will monitor gas concentrations and temperature gradients, which will help to explain how animals achieve suitable ventilation underground and maintain body temperature. To investigate animal behaviour, tracking collars will be equipped with miniature sensors, such as accelerometers and magnetometers, which will record motion and energetics. To reduce data volumes, tracking collars will automatically characterize animal behaviour primitives, such as walking or sleeping. To further increase the rate of "learning" this information, tracking collars will share motion features, forming a distributed knowledge base. Thus, this research proposes a broad animal monitoring and tracking system, which will reveal a complete picture of animal life underground, for the first time.<br /><br />To achieve the goals of the research a close collaboration with the Wildlife Conservation Research Unit at the University of Oxford will be formed. They will guide the design of the tracking collars and attach them to suitable badgers during regular research undertaken in Wytham Woods, Oxfordshire. Their expertise is also vital in framing the research to address biologically relevant questions. Data from this system will also be used by researchers in the University of Cambridge Computing Laboratory, to investigate social contact networks. Ultimately this insight into the detail of badgers' lives will help to unravel the true extent with which they interact with each other, and shed light not just on the behavioural-ecology of this species, but investigate their social systems and address important questions concerning the transmission of disease.</p> Department of Computer Science, University of Oxford Projects All oucl-All-UnderTracker Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/CPROVER/ <p>The cost of software quality assurance (QA) dominates the cost of IT development and maintenance projects. QA is frequently on the critical path to market. Effective software QA is therefore decisive for the competitiveness of numerous industries that rely on IT, and essential for government tasks that rely heavily on IT.</p> <p>This research programme will provide a pragmatic solution to the most pressing issue in software QA in mainstream software engineering: the use of concurrency. Programmers make use of numerous favors of concurrency in order to achieve better scalability, savings in power, increase reliability, and to boost performance. The need for software that makes diligent use of concurrent computational resources has been exacerbated by power-efficient multi-core CPUs, which are now widely deployed, but still unfertilized due the lack of appropriate software. Concurrent software is particularly difficult to test, as bugs depend on particular interlavings between the sequential computations. Defects are therefore difficult to reproduce and diagnose, and often elude even very experienced programmers.</p> <p>We propose to develop new, ground-braking reasoning and testing technology for this kind of software, with the goal of cutting the staff effort in QA of concurrent effort in half. We will use a tightly integrated combination of scalable and performant testing technology and Model Checking and abstract interpretation engines to prune the search. Every aspect of the research programme is geared towards improving the productivity of the average application programmer. Our theories and reasoning technology will therefore be implemented in a seamless fashion within the existing, well-accepted programming environments Visual Studio and Eclipse, in close collaboration with Microsoft and IBM.</p> Department of Computer Science, University of Oxford Projects All oucl-All-CPROVER Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/FSCM/ <p>Words are the building blocks of sentences, yet meaning of a sentence goes well beyond meanings of the words therein. Indeed, while we do have dictionaries for words, we don't seem to need them to infer the meaning of a sentence from meanings of its constituents. Discovering the process of meaning assignment in natural languages is one of the most foundational issues in linguistics and computer science, whose findings will increase our understanding of cognition and intelligence and may assist in applications to automating language-related tasks, such as document search as done by Google.</p> <p>To date, the compositional logical and the distributional probabilistic models have provided two complementary partial solutions to the problem of meaning assigning in natural languages. The logical approach is based on classic ideas from mathematical logic, mainly Frege's principle that meaning of a sentence can be derived from the relations of the words in it. The distributional model is more recent, it can be related to Wittgenstein's philosophy of 'meaning as use', whereby meanings of words can be determined from their context. The logical models have been the champions on the theory side, whereas in practice their probabilistic rivals have provided the best predictions. This two-sortedness of defining properties of meaning: 'logical form' versus 'contextual use', has left the question of 'what is the foundational structure of meaning?' even more open a question than before. This project has ambitious and far-reaching goals; it aims to bring together these two complementary concepts to tackle the question. And it aims to do so by bridging the fields of linguistics, computer science, logic, probability theory, category theory, and even physics. Its scope is foundational, multi and inter disciplinary, with an eye towards applications.</p> <p>Meaning assignment is a dynamic interactive process involving grammar and logic as well as meanings of words. Both of the two existing approaches to language miss a crucial aspect of this process: the logical model ignores meanings of words, the distributional model ignores the grammar and logic. We aim to model the entire dynamic process alongside the following three strands of integration, foundations, and applications.</p> <p>(I) In integration we develop a process of meaning assignment that acts with the compositional forms of the logical model on the contextual word-meaning entities of the distributional model.</p> <p>(II) In foundations, we go beyond classical logical principles of compositionality and context-based models of meaning to develop more fundamental processes of meaning assignments based on novel information-flow techniques, mainly from physics, but also from other linguistic approaches and other models of word meaning, such as ontological domains and conceptual spaces.</p> <p>(III) In applications, we evaluate our theories against naturally occurring data and apply the results to practical issues based on meaning inference and similarity, e.g. in search. To be able to work with logical connectives in Google, one needs to re-enter them by hand in the 'advanced search' tab, by manually decomposing the logical structure of the sentence and moreover providing the extra context for their different meanings. This is fundamentally non-compositional and goes against the spirit of automated search. It is exactly here that the lack of compositional methods in meaning assignment causes practical problems and where our compositional methods become of use. Hence, we aim to put forward our results to tackle such problems, e.g. to be able to use our sentence similarity models for paraphrasing, question-answering, and retrieving documents that have the same meaning and/or are about the same subject. Our proposed partnership with Google, ensures access to real life data and helps implementation and applicability of our methods in small and large scales.</p> Department of Computer Science, University of Oxford Projects All oucl-All-FSCM Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/PINCETTE/ <p>The vision of PINCETTE is to solve the problem of high cost of changes by introducing an automated framework and methodology, and a mix of technologies to identify the impact of changes that derive from intra-component changes (due to error fixing and functionality enhancement) and from component replacement within a single product and a product family. This methodology improves the reliability of networked software by implementing an innovative solution for the automatic detection, localization, and repair of program bugs.</p> Department of Computer Science, University of Oxford Projects All oucl-All-PINCETTE Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/sensing/ <p>Wireless sensor networks have a multitude of applications, for example in environmental monitoring, autonomous transport, or intelligent buildings. There are significant challenges in engineering and programming of effective sensor-based systems, in view of resource limitations of the devices and unreliability of wireless links. At the same time, sensor networks are increasingly often deployed in safety‐critical contexts that require sound and autonomous decision making and guarantees on their correct behaviour. This research will bring together two hitherto separate streams of work: (1) Trigoni&rsquo;s group, focused on developing systems and novel algorithmic techniques for data collection in sensor networks, and (2) Kwiatkowska&rsquo;s VERIWARE project, which aims to develop pioneering methodology and software tools to ensure the reliability and correct behaviour for such devices, thus enabling testbed experimentation and evaluation of the techniques that is not feasible as part of existing projects.</p> Department of Computer Science, University of Oxford Projects All oucl-All-sensing Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/synergy/ <p>Synergy will develop a simulation environment and a decision-support system aiming at enabling deployment<br />of systems medicine. The three core elements are a knowledge base (KB), an inference engine (IE), and a<br />graphical visualisation environment (GVE). The project focuses on patients with chronic obstructive pulmonary<br />disease (COPD).<br />The KB will include five well established physiological models addressing: 1) Central and peripheral O2 transport<br />and utilization, 2) Pulmonary gas exchange, 3) Regional-lung heterogeneities in ventilation and perfusion, 4)<br />Skeletal muscle bioenergetics, and 5) Mitochondrial reactive oxygen species (ROS) generation. These models<br />will be written in systems biology markup language (SBML) and vertically integrated. Ontologies will be used<br />as the default knowledge-representation system. The KB will include multi-level data from experimental studies<br />(BioBridge), data from a multicentre longitudinal study on COPD phenotyping (PAC-COPD) and public datasets.<br />The IE will enable to explore associations over the KB, perform transversal multi-scale model integration and<br />related simulations including interactions among O2-availability/O2-utilization, ROS generation, systemic<br />inflammation and abnormal tissue remodelling.<br />The Web-based GVE will facilitate relevant simulations in a more intuitive way with respect to the state of the art,<br />addressing two main user profiles: bio-researchers and clinicians.<br />The focus will be on underlying mechanisms of COPD phenotypes associated with poor prognosis. Disease<br />model validation and refinement will be done using a well-established, large dataset (ECLIPSE) together with<br />experimental studies designed to test &ldquo;in silico&rdquo; generated hypotheses. Besides the use of the simulation<br />environment by bio-researchers for optimal experimental design, the Synergy platform will be a relevant<br />decision-support tool for integrated healthcare strategies aiming at modulating the evolution of COPDs.</p> Department of Computer Science, University of Oxford Projects All oucl-All-synergy Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/AirPROM/ <p>The airways diseases asthma and chronic obstructive pulmonary disease affect over 400 million people world-wide and cause considerable morbidity and mortality. Airways disease costs the European Union in excess of &euro;56 billion per annum. Current therapies are inadequate and we do not have sufficient tools to predict disease progression or response to current or future therapies. Our consortium, Airway Disease PRedicting Outcomes through Patient Specific Computational Modelling (AirPROM), brings together the exisiting clinical consortia (EvA FP7, U-BIOPRED IMI and BTS Severe Asthma), and expertise in physiology, radiology, image analysis, bioengineering, data harmonization, data security and ethics, computational modeling and systems biology. We shall develop an integrated multi-scale model building upon existing models.</p> <p>This airway model will be comprised of an integrated 'micro-scale' and 'macro-scale' airway model informed and validated by omic data and ex vivo models at the genome-transcriptome-cell-tissue scale and by CT and functional MRI imaging coupled to detailed physiology at the tissue-organ scale utilising Europe&rsquo;s largest airway disease cohort. Validation will be undertaken cross-sectionally, following interventions and after longitudinal follow-up to incorporate both spatial and temporal dimensions.</p> <p>AirPROM has a comprehensive data management platform and a well-developed ethico-legal framework. Critically, AirPROM has an extensive exploitation plan, involving at its inception and throughout its evolution those that will 'develop' and 'use' the technologies emerging from this project. AirPROM therefore will bridge the critical gaps in our clinical management of airways disease, by providing validated models to predict disease progression and response to treatment and the platform to translate these patient-specific tools, so as to pave the way to improved, personalised management of airways disease.</p> Department of Computer Science, University of Oxford Projects All oucl-All-AirPROM Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/QueRe/ <div><strong>Background and Motivation:</strong></div><div>&nbsp;</div><div>Critical decisions in industry, science, government, and healthcareare are increasingly based on information derived from data sources whose number, size and heterogeneity continue to grow at a phenomenal rate. Improved access to and exploitation of such data is thus becoming critical to both our future competitiveness and our quality of life. One such example in the telecoms industry relates to the organisation and management of global communications networks, which has become highly complex, and involves the adjustment in real time of numerous parameters. Decisions on how to adjust these parameters can have a huge impact on performance and on profitability. These decisions are based on information derived from a large number of data sources, ranging from historical data to data streaming from monitoring equipment.</div><div><br /></div><div>In order to meet the challenges presented by this kind of application, a new generation of Ontology-Based Information Systems (OBISs) is beginning to emerge. These systems aim to offer high performance and large storage capacities, while at the same time exploiting rich knowledge about the application domain captured in ontologies; this knowledge is used to present a unified view over the data sources, provide a domain-centric vocabulary for use in queries, deal with incomplete and semi-structured data, and enrich query answers with implicit information. &nbsp;The hope is that OBISs can be realised via a principled synthesis of ontological reasoning and database management techniques in a flexible architecture that will enable them to access data directly from different kinds of data repository, exploit different kinds of query answering technology, and adapt to a wide range of different application settings, while still providing formal guarantees about the quality of query answers.</div><div><br /></div><div>Currently, all practical approaches to building OBISs rely on query answering via query rewriting: answers to a query q over an ontology O and a database instance DB are computed by first rewriting q (using O) into another query q' and then evaluating q' over DB. Query rewriting means that query evaluation can be delegated to existing information systems and thus exploit the capabilities that they provide, including robust scalability and the ability to handle rapidly changing data. However, existing query rewriting techniques have focused on reducing the expressivity of the ontology language in order to obtain formal tractability guarantees; this seriously restricts the modelling capabilities of the system, while not necessarily delivering robust scalability in practice.</div><div><br /></div><div>The goal of the project is to design practically effective query answering algorithms for more expressive ontology languages with features such as transitivity and equality, and demonstrating that OBISs based on such languages can be effectively deployed in applications. In achieving these goals we will work closely with&nbsp;<a href="#mce_temp_url#">Alcatel-Lucent Bell Labs</a>, who have extensive technical expertise in this area, as well as motivating applications in the telecoms domain.</div><div>&nbsp;</div><div><strong>Technical Details:</strong></div><div>&nbsp;</div><div>Due to the fundamental trade-off between ontology language expressivity and query answering complexity, the currently prevalent view is that Ontology-Based Information Systems (OBISs) should use ontology languages that allow for first-order reducibility of query answering, with the low data complexity of the resulting algorithm often being given as a supporting argument. Reasoning with individual equality, transitivity, recursive axioms, or any kind of disjunctive information is LogSpace-hard in data complexity, which prevents first-order reducibility, so these features were excluded from OBIS languages such as OWL 2 QL. Such features, however, are often needed in practice; for example, transitivity is needed to describe part-whole relationships, and equality is critical in information integration. &nbsp;Furthermore, even with very restricted ontology languages, the existing query rewriting algorithms produce rewritings whose size is of the order ((|O| . |q|)^|q|), where |q| and |O| are the sizes of the original query and the ontology, respectively. Data complexity is useful only when the size of the query is negligible w.r.t. the size of the data; but then, since the rewritings can be large, the low data complexity of query answering via rewriting does not provide a realistic estimate of performance in practice. Thus, while seriously restricting the ontology language, first-order reducibility does not provide a practical scalability guarantee.</div><div><br /></div><div>The technical challenge of this project is to develop a new approach to query answering in OBISs. On the one hand, in order to accommodate the expressivity required in practice, more expressive languages should be considered as targets for query rewriting. On the other hand, finer-grained complexity measures are needed in order to identify assumptions about the content and usage of the OBIS under which scalability can still be guaranteed.&nbsp;</div><div><br /></div><div>Regarding more expressive targets for query rewriting, first steps in this direction have already been taken: if an ontology O is expressed in the DL ELHIO, a conjunctive query q over O can be rewritten into a Datalog query q'. &nbsp;The polynomial data complexity of Datalog may be too high in general; however, little is known about using languages with data complexity between AC0 and PTime as targets for query rewriting. Regular path queries (RPQs) are a promising candidate since they provide a limited form of recursion. RPQs were used as the basis for the query languages in semi-structured databases, so a rewriting approach based on RPQs should allow for ontology-based access to semi-structured data.</div><div><br /></div><div>Regarding finer-grained complexity measures, in related fields such as propositional satisfiability, constraint satisfaction and conjunctive query answering, measures based on (hyper)treewidth have proved very effective. Adapting such measures to OBISs, and developing query answering algorithms that exploit them, is thus an important research problem. It is currently unknown whether assumptions about the structure of data (rather than on the structure of the ontology and/or the query) may be used to obtain practically effective query answering algorithms. Furthermore, determining classes of ontology languages and/or queries that admit `small' rewritings is an important and practically relevant open problem.</div><div><br /></div><div>Some usage scenarios may also allow alternative architectures for OBISs to be considered. &nbsp;For example, if the ontology and the data change relatively infrequently, ontology-based ISs could be based on the &quot;combined&quot; query answering techniques, where some entailed facts are materialised in the data. It is, however, currently unclear whether these can be extended to expressive languages such as Horn-SHIQ and the related dependency classes. Furthermore, if the data changes more frequently, the cost of recomputing the preprocessed database instance may be prohibitive; in such cases, however, an incremental maintenance approach may be efficacious.</div><div>&nbsp;</div><div><br /></div> Department of Computer Science, University of Oxford Projects All oucl-All-QueRe Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/TCLOUDS/ <P>Protecting critical infrastructures providing communications, energy, or healthcare presents increasing ICT challenges as ICT itself has become vital to them.</P> <P><BR>Internet-scale ICT infrastructures (“Infrastructure Clouds”) promise scalable virtualised computing, network, and storage resources over the Internet. They provide scalability and cost-efficiency but pose significant new privacy and resilience challenges.</P> <P><BR>Clouds may evolve into a single point of failure, threaten all dependent ICT, and put the Future Internet at risk.</P> <P><BR>TCLOUDS builds a resilient Future Internet platform by progress in four areas:</P> <UL> <LI>Addressing the legal and business implications while building a regulatory framework for enabling privacy-enhanced cross-border infrastructure clouds.</LI> </UL> <UL> <LI>Architecture and prototypes for a federation of trustworthy infrastructure clouds that build on complementary and mutually re-enforcing technical approaches:</LI> </UL> <BLOCKQUOTE> <OL> <LI>A Trustworthy Infrastructure Cloud enables individual providers to offer more resilient and privacy-aware infrastructure clouds accessible via open interfaces.</LI> <LI>Privacy and Resilience for Commodity Clouds enables end users to put a security layer on top of existing commodity infrastructure clouds to enforce their security objectives. This enables integration of commodity clouds into the TCLOUDS federation.</LI> <LI>Federated Cloud-of-cloud Middleware offers privacy-protection and resilience beyond any individual cloud. This expands trust from trusted (enterprise-internal) clouds to less trusted (off-shored) ones, or federates a set of partially trusted providers into a trustworthy and adaptive federation that furthermore prevents lock-in to a given dominating offering.</LI> </OL> </BLOCKQUOTE> <UL> <LI>Validation and impact through cloud scenarios:</LI> </UL> <BLOCKQUOTE> <OL> <LI>Smart power grids connect renewable energy sources and users. It is a premier example of an Internet of Things.</LI> <LI>Home healthcare provides prophylaxis to citizens. We focus on the privacy and usability challenges of cross-border usage of personal data.<BR>Collaboration with complementary standardisation and FP7 projects maximises impact and fosters an Open European Trustworthy Cloud ecosystem.</LI> </OL> </BLOCKQUOTE> <FONT size="2">Oxford's involvement  is chiefly in the scientific activity of the project, developing previously sketched research architectures into viable detailed designs, and undertaking analysis of these. Furthermore UOXF will be in charge of end-to-end security for applications as well as job security manager for controlled data access. The University will play an important role in dissemination through scientific publications and conferences, and continue to play host to conferences and summer schools in relevant topics.</FONT> Department of Computer Science, University of Oxford Projects All oucl-All-TCLOUDS Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/webinos/ <p>webinos is an EU-funded project aiming to deliver a platform for web applications across mobile, PC, home media (TV) and in-car devices.</p> <p>&nbsp;</p> <p>The webinos project will define and deliver an Open Source Platform and software components for the Future Internet in the form of web runtime extensions, to enable web applications and services to be used and shared consistently and securely over a broad spectrum of converged and connected devices, including mobile, PC, home media (TV) and in-car units.</p><p>&nbsp;</p><p><strong>webinos in a Nutshell: </strong>Promoting a &ldquo;single service for every device&rdquo; vision, webinos will move the existing baseline of web development from installed applications to services, running consistently across a wide range of connected devices, ensuring that the technologies for describing, negotiating, securing, utilizing device functionalities and adapting to context are fit for purpose.</p> <p>Innovations in contextual description will be broad covering but not limited to device capabilities, network access, user identity and preferences, location, behaviourally induced properties and finally the more complex issue of the users&rsquo; social network context and social media engagement.</p> <p>webinos will boost the industry migration towards web-based services. webinos can back this by providing inter-operable, standardised, open source technology utilizable across domains with direct commercially exploitable value. webinos will also act as an industry catalyst to encourage collaboration and discourage fragmentation in this space. There are strong industry moves towards Internet friendly and Internet integrated offerings, and there exists a window of opportunity to place the webinos technology on a robust open foundation that will remove economic barriers to engagement, embody policy on data privacy in concrete technology and creating a centre of web centric expertise.</p><p>&nbsp;</p><p>(webinos was initially known as WAX).&nbsp;</p> Department of Computer Science, University of Oxford Projects All oucl-All-webinos Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/VERIWARE/ In the words of Adam Greenfield, &ldquo;the age of ubiquitous computing is here: a computing without computers, where information processing has diffused into everyday life, and virtually disappeared from view&rdquo;. Conventional hardware and software has evolved into &lsquo;everyware&rsquo; &ndash; sensor-enabled electronic devices, virtually invisible and wirelessly connected &ndash; on which we increasingly often rely for everyday activities and access to services such as banking and healthcare. The key component of &lsquo;everyware&rsquo; is software, embedded inside electronic gadgets and continuously interacting with its environment by means of sensors and actuators. Ubiquitous computing must deal with the challenges posed by the complex scenario of communities of &lsquo;everyware&rsquo;, in presence of environmental uncertainty and resource limitations, while at the same time aiming to meet high-level expectations of autonomous operation, predictability and robustness. This calls for the use of stochastic modelling, discrete and continuous dynamics, quantitative measures, and goal-driven approaches, which the emerging quantitative software verification is unable to address at present. <p> The central premise of the VERIWARE project is that there is a need for a paradigm shift in verification to enable &lsquo;everyware&rsquo; verification, which can be achieved through a model-based approach that admits discrete and continuous dynamics, the replacement of offline methods with online techniques such as machine learning, and the use of game-theoretic and planning techniques. The project will significantly advance quantitative probabilistic verification in new and previously unexplored directions. This will involve investigating the fundamental principles of &lsquo;everyware&rsquo; verification, development of algorithms and prototype implementations, and experimenting with case studies.</p><p>&nbsp;</p> Department of Computer Science, University of Oxford Projects All oucl-All-VERIWARE Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/DIADEM/ <p>Do you need to rent a new apartment? Or would you just like to find a restaurant in your area that serves "pasta al pesto" as today&rsquo;s special? In either case, you would most likely start a web search. But keyword search is not really appropriate in such cases, because you risk being swamped with irrelevant information, rather than finding what you want. If all the information were available in structured form, you could find what you are looking for much faster. Search engine providers such as Google, Yahoo! and Microsoft are aware of this, and are keenly looking for new methods that automatically recognize and extract data from domain-specific websites with semi-structured content. To this date, this problem has not been satisfactorily solved; its solution seems to require a major research breakthrough.</p> <p><br />In this project we will tackle precisely this challenge. Our goal is very ambitious. We want to develop domainspecific data extraction systems that take as input a URL of a website in a particular application domain, automatically explore the web site, and deliver as output a structured data set containing all the relevant information present on that site. We will provide the logical, algorithmic, and methodological foundations for the knowledge-based extraction of structured data from web sites belonging to specific domains, and we will develop two extraction systems for two different domains. To achieve our goal, we will design new methods and algorithms that combine database techniques with methods of knowledge representation and reasoning and web data extraction techniques. The breakthrough in automatic data extraction, which we are striving for, would enable a leap forward for two interrelated technologies which are the hottest emerging topics in web search: vertical search, that is, web search in specialized domains, and object search, that is, the search for web data objects rather than web pages.</p> <p>For more details and results see the <a href="http://diadem.cs.ox.ac.uk/">DIADEM homepage</a>.&nbsp;</p> Department of Computer Science, University of Oxford Projects All oucl-All-DIADEM Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/ComputertoClinic/ <span>Heart Failure (HF) is defined by the heart's reduced ability to pump blood due to a drop in cellular contractility, enlarged anatomy and increased coronary micro-vascular resistance. This loss of pump function accounts for a significant increase in both mortality and morbidity in western society. With the U.K.'s elderly population expanding, HF is rapidly becoming an epidemic. There is currently a 1 in 5 life-time risk of HF and costs associated with acute and long term hospital treatments are accelerating. The significance of the disease has motivated the application of state of the art clinical imaging techniques to aid diagnosis and clinical planning. Measurements of cardiac wall motion, chamber flow patterns and coronary perfusion currently provide high resolution data sets for characterising HF patients. However, the clinical practice of using population-based metrics derived from separate image sets often indicates contradictory treatments plans due to inter-individual variability in pathophysiology. Thus, despite imaging advances, determining optimal treatment strategies for HF patients remains problematic. To exploit the full value of imaging technologies, and the combined information content they produce, requires the ability to integrate multiple types of functional data into a consistent framework. This in turn will support a paradigm shift away from predefined clinical indices determining treatment options and a move towards true personalisation of care based on an individual's physiology.<br /><br />An exciting and highly promising strategy for underpinning this shift is the assimilation of multiple image sets into personalised and biophysically consistent mathematical models. The development of such models provides the ability to capture the multi-factorial cause and effect relationships which link the underlying pathophysiological mechanisms. Furthermore, using a biophysical basis presents unique opportunities to assist with treatment decisions through the derivation of quantities that cannot be imaged but are likely to play a key mechanistic role in HF e.g. tissue stress and pump efficiency.<br /><br />In parallel with imaging advances the approach is also underpinned by the ongoing development of complementary technologies, including improved numerical methods and increased performance per unit cost of computing. This computational progress has accelerated the addition of multi-physics functionality to a range of organ models which have recently been organized into international initiatives such as the IUPS sponsored Physiome and VPH projects. Within these programmes the heart is arguably the most advanced current exemplar of an integrated organ model. As such it represents a promising first candidate with which to focus on an important human disease.<br /><br />My goal during this fellowship will be to focus on personalising and applying these models in clinical and industrial settings for treating HF patients. Model simulations will be focused on quantifying diagnosis, aiding patient selection and guiding interventional planning for specific treatments carried out by leading clinicians based in the cardio-vascular imaging group at Kings College London (KCL). In addition to this direct clinical application of the model, the research will also be focused on the tuning of Left Ventricular Assist Devices (LVADs) which are often connected to the heart in HF to reduce mechanical load by pumping blood from the left ventricle directly into the aorta. <br /><br />Through these applications my aim is to both improve our understanding of this significant cardiovascular disease and demonstrate the potential of biophysical models for improving human healthcare.</span> Department of Computer Science, University of Oxford Projects All oucl-All-Computer to Clinic Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/VerificationofShared-MemoryConcurrentSoftware/ <div>The goal of this project is to develop algorithms and tools for&nbsp;automated formal verification of low-level C programs that make use of&nbsp;shared-variable concurrency. &nbsp;In spite of extensive research on&nbsp;concurrent computation, almost all existing program analysis tools are&nbsp;limited to sequential programs or programs that communicate via some&nbsp;form of explicit message passing. However, shared-variable concurrency&nbsp;is the predominant form of concurrency in commercial environments, and&nbsp;tool support is in dire need. The project focuses on</div><div><ol><li>verification&nbsp;by means of automated summarisation of threads,</li><li>identification of&nbsp;transactions, enabling partial-order reductions, and</li><li>Craig&nbsp;interpolation to derive thread invariants.</li></ol></div><div><br /></div> Department of Computer Science, University of Oxford Projects All oucl-All-Verification of Shared-Memory Concurrent Software Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/QuantitativeVerification/ <p>Enabling engineers, programmers, and researchers to automatically verify the correctness of the computer systems that they design is one of the Grand Challenges of computing research. The scientific and economic importance of this goal has long been recognised, yet, despite substantial progress in basic and applied research over the last few decades, much work remains ahead of us. With this Fellowship I would like to address the issue of quantitative verification and analysis of real-time and probabilistic systems, from the development of novel and fundamental algorithms all the way to the design and implementation of tools.</p><p> The field of computer-aided formal verification is concerned, broadly speaking, with ensuring that computer systems function as they were intended to. In any instances, the correctness of such systems is critical, with failures potentially incurring disastrous financial consequences or even loss of human life. Well-known catastrophic failures in recent years include the 1996 explosion of the Ariane 5 rocket on its maiden flight, the 1999 loss of the MarsPolarLander as it attempted touch-down, and the 2003 Northeast power blackout, costing several billion dollars. In each case, the culprit was a bug in the software, which investigators subsequently discovered could have been rooted out by using computer-aided verification technology.</p><p> One of the key approaches to verification is model checking, pioneered nearly three decades ago by Clarke, Emerson, Queille, Sifakis, and Pnueli, for which Clarke, Emerson, and Sifakis won the 2007 Turing Award and Pnueli the 1996 Turing Award. In this framework, a finite-state model of the system under consideration is constructed, usually in an automated or semi-automated manner, and its specifications are expressed using formulas in a suitable temporal logic. Model checking then involves determining whether the model satisfies the formulas, which is achieved using a sophisticated mix of abstraction techniques, symbolic representations, and graph-theoretic algorithms. The fundamental challenge in this approach is the state space explosion problem, whereby the state space of the systems to be explored can easily turn out to be astronomical. Model checking has nonetheless proved hugely valuable to the hardware industry, and more recently has become increasingly successful in software development. Industrial users of model checking include Intel, Motorola, Microsoft, Airbus, NASA, and many others.</p><p> While traditional model checking has focussed on qualitative representations and properties of systems, there is an urgent need for quantitative verification technology, and specifically for frameworks able to accurately handle real-time and probabilistic systems. Such systems are ubiquitous, occurring in cars, aeroplanes, railway controllers, power plants, medical equipment, the Internet, etc. Research in quantitative verification has made great strides over the last two decades, yet many fundamental challenges remain. Building on recent breakthroughs in real-time and probabilistic model checking, I believe we are now ideally poised to tackle many of these challenges.</p> Department of Computer Science, University of Oxford Projects All oucl-All-Quantitative Verification Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/GAMES/ <p>As computing systems become larger, more complex, and increasingly distributed and interactive, there is a pressing need for formal methods that guarantee their reliability, correctness, and efficiency. This network proposes a research and training programme for the design and verification of computing systems, using a methodological framework that is based on the interplay of finite and infinite games, mathematical logic and automata theory.</p> Department of Computer Science, University of Oxford Projects All oucl-All-GAMES Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/accelerating/ The aim of this project is to increase the rate of progress in clinical cancer research by accelerating the development, running, and re-analysis of translational early-phase studies through the provision of semantics-driven software solutions. It will take an approach that has proven effective for randomised controlled trials, where therapies are tested on large populations, and apply it to the earlier stages of research and development, where the effect of therapies upon humans are first explored, and significant discoveries are made. It will build upon the work of the UK CancerGrid project, which has developed techniques for the provision of software systems in which data is associated with a detailed, computable representation of its semantics, and that association is used to drive subsequent processing &ndash; analysis, integration, and re-use. It will employ Microsoft technologies ideally positioned for the implementation of these techniques: in particular, SharePoint and Silverlight.&nbsp; It will complement workflow and community-building technologies under development at Microsoft, as part of a platform for leading-edge, collaborative science. Department of Computer Science, University of Oxford Projects All oucl-All-accelerating Fri, 07 Dec 2012 02:50:39 GMT http://www.cs.ox.ac.uk/projects/SPROUT/ <P> <IMG title="SPROUT logo" src="/people/Dan.Olteanu/sprout_logo-small.png" alt="SPROUT logo" align="left"> </P> <P>Today, uncertainty is commonplace in data management scenarios dealing with data integration, sensor readings, information extraction from unstructured sources, and whenever information is manually entered and therefore prone to inaccuracy or partiality. Key challenges in probabilistic data management are to design probabilistic database formalisms that can compactly represent large sets of possible interpretations of uncertain data together with their probability distributions, and to efficiently evaluate queries on very large probabilistic data. Such queries could ask for confidences in data patterns possibly in the presence of additional evidence. The problem of query evaluation in probabilistic databases is still in its infancy. Little is known about which queries can be evaluated in polynomial time, and the few existing evaluation methods employ expensive main-memory algorithms.</P> <P>The aim of this project is to develop techniques for scalable query processing in probabilistic databases and use them to build a robust query engine called <STRONG> <SPAN style="color: #008000;">SPROUT</SPAN> </STRONG> ( <SPAN style="color: #008000;"> <STRONG>S</STRONG> </SPAN> calable <SPAN style="color: #008000;"> <STRONG>PRO</STRONG> </SPAN> cessing on <SPAN style="color: #008000;"> <STRONG>U</STRONG> </SPAN> ncertain <SPAN style="color: #008000;"> <STRONG>T</STRONG> </SPAN> ables). We are currently exploring three main research directions.</P> <UL> <LI>We are investigating open problems in efficient query evaluation. In particular, we aim at discovering classes of tractable (i.e., computable in polynomial time wrt data complexity) queries on probabilistic databases. The query language under investigation is SQL (and its formal core, relational algebra) extended with uncertainty-aware query constructs to create probabilistic data under various probabilistic data models (such as tuple-independent databases, block-independent disjoint databases, or U-relations of MayBMS). </LI> <LI>For the case of intractable queries, we investigate approximate query evaluation. In contrast to exact evaluation, which computes query answers together with their exact confidences, approximate evaluation computes the query answers with approximate confidences. We are working on new techniques for approximate query evaluation that are aware of the query and the input probabilistic database model (tuple-independent, block-independent disjoint, etc). </LI> <LI>Our open-source query engine for probabilistic data management systems uses the insights gained from the first two directions. This engine is based on efficient secondary-storage exact and approximate evaluation algorithms for arbitrary queries. </LI> </UL> <H3>News</H3> <UL> <LI>[Sept 2012] Quentin Spencer-Harper was awarded the Gloucester Research Project Prize for best 3rd year CS project 2012. His project investigates the evaluation problem of queries with aggregates on probabilistic databases. </LI> </UL> <UL> <LI>[Feb 2012] Robert Fink will do an internship over the summer of 2012 in the research group of Alon Halevy at Google Inc (Mountain View).</LI> </UL> <UL> <LI>[Sept 2011] Larisa Han won the Hoare prize for the best MSc thesis in Computer Science at Oxford for her thesis on aggregates in probabilistic databases.</LI> </UL> <UL> <LI>[July 1, 2011] A news item about our work on probabilistic databases and the new book has been featured on <A href=" http://www.i-programmer.info/news/84-database/2653-probabilistic-databases-the-next-big-thing.html">i-programmer </A>. Thank you! </LI> </UL> <UL> <LI>[June 2, 2011] The first book on the foundations of <A href="http://www.morganclaypool.com/doi/abs/10.2200/S00362ED1V01Y201105DTM016">Probabilistic Databases</A> by Dan Suciu, Dan Olteanu, Christopher Re, and Christoph Koch has been published by Morgan & Claypool!</LI> </UL> <UL> <LI>[June 1, 2011] An article on probabilistic databases authored by Robert and Dan appeared in the 10th anniversary edition of <A href="http://www.thinkdigit.com/">The Digit Magazine</A>, the largest selling technology magazine in India, with a certified readership of 200,000.</LI> </UL> <UL> <LI>[June 1, 2011] An article featuring SPROUT has appeared in the second edition of Inspired Research, the industry letter of the Department of Computer Science at Oxford.</LI> </UL> <UL> <LI>[Oct 2010] Swaroop Rath's MSc thesis on the evaluation of full relational algebra queries was awarded distinction. His overall performance placed him second in the MSc in CS course.</LI> </UL> <UL> <LI>[Oct 2009] Rasmus Wissmann won the Hoare prize for the best overall performance in our MSc in Computer Science at Oxford. His thesis on "Tractable Queries with Inequalities on Probabilistic Databases" was also awarded distinction.</LI> </UL> <UL> <LI>[July 2009] Our SIGMOD'09 paper on tractable conjunctive queries with inequalities passed the SIGMOD'09 repeatability and workability evaluation (RWE). Out of 63 accepted papers, 19 participated in RWE, and 10 papers passed RWE.</LI> </UL> <UL> <LI>[May 2009] Jiewen Huang won a Travel Grant from ACM SIGMOD to attend SIGMOD 2009! Read more <A href="http://www.sigmod09.org/delegates_travel_grants.shtml">here</A> about the grant. </LI> </UL> <UL> <LI>[Sept 2008] Jiewen Huang won the Hoare prize for the best MSc thesis in Computer Science at Oxford. His thesis on "Scalable Query Evaluation for Tuple-Independent Probabilistic Databases" investigates aspects of this project and the results are reported in the SUM'08 and ICDE'09 papers. <BR> Jiewen also won the Hoare prize for the best overall performance in his MSc year at Oxford. </LI> </UL> <UL> <LI>[August 2008] SPROUT prototype released as part of the MayBMS database management system available at <A href="http://maybms.sourceforge.net/">maybms.sourceforge.net</A></LI> </UL> <H3>System prototype</H3> <P>SPROUT is implemented as an extension of the PostgreSQL backend with secondary-storage and main-memory algorithms for exact or approximate confidence computation. The latest online version of SPROUT is available <A href="/people/dan.olteanu/code/ohk-icde10-sprout.tgz">here</A> (this is the version reported in the ICDE 2010 paper; newer versions are available on request). Major releases of SPROUT are also included in the MayBMS database management system available at <A href="http://maybms.sourceforge.net/">maybms.sourceforge.net</A>. The latest MayBMS version (that can be checked out from the sourceforge svn) uses the SPROUT version reported in the ICDE 2009 paper.</P> <H3>Talks</H3> <UL> <LI><EM>Aggregation in Probabilistic Databases</EM> [<A href="/people/Dan.Olteanu/tutorials/vldb12-talk.pdf">PDF</A>]<BR> Given by Robert at various venues in 2011-2012.</LI> </UL> <UL> <LI><EM>On the Optimal Approximation of Queries using Tractable Propositional Theories</EM>. [<A href="/people/Dan.Olteanu/tutorials/icdt11-talk.pdf">PDF</A>]<BR>Given by Robert at various venues in 2011-2012. </LI> </UL> <UL> <LI><EM>The Case of Relational Algebra Queries in Probabilistic Databases</EM><BR> Given by Dan at IC Colloquium, EPFL, January 2011, Lausanne.</LI> </UL> <UL> <LI><EM>A Look at Probabilistic Databases through SPROUT glasses</EM><BR> Given by Dan at the Statistics Departmental Seminar, November 2010, Oxford.</LI> </UL> <UL> <LI><EM>SPROUT: Scalable Query Processing in Probabilistic Databases</EM> [Feb 2010 overview talk <A href="/people/Dan.Olteanu/tutorials/mound10-talk.pdf">PDF</A>, <A href="/people/Dan.Olteanu/tutorials/sprout-poster.pdf">A1 poster.pdf</A>]<BR> Given by Dan at various venues, 2008-2009. </LI> </UL> <UL> <LI><EM>A Toolbox of Query Evaluation Techniques for Probabilistic Databases.</EM> [<A href="/people/dan.olteanu/tutorials/mound10-talk.pdf">PDF</A>]<BR> Invited talk at the Workshop on Logic in Databases (LID), October 2009. <BR> Invited talk at the Workshop on Management and mining Of UNcertain Data (MOUND), In conjunction with ICDE, March 2010. </LI> </UL> <UL> <LI><EM>Incomplete and Probabilistic Data Management.</EM> 3-hour tutorial given at FOX training week, October 2009. </LI> </UL> <H3>Theses</H3> <UL> <LI>Lampros Papageorgiou: <EM>Pigora: An Integration System for Probabilistic Data</EM><BR> MSc in CS, Oxford 2012. </LI> </UL> <UL> <LI>Andrei Marin: <EM>Incremental View Maintenance in Probabilistic Databases</EM><BR> MSc in CS, Oxford 2012. </LI> </UL> <UL> <LI>Quentin Spencer-Harper: <EM>Evaluation of Aggregate Queries on pvc-Tables</EM><BR> BSc in CS, Oxford 2012. </LI> </UL> <UL> <LI>Larisa Han: <EM>Aggregates in Probabilistic Databases</EM><BR> MSc in CS, Oxford 2011. </LI> </UL> <UL> <LI>Swaroop Rath: <EM>Efficient Evaluation of Full Relational Algebra on Probabilistic Databases</EM><BR> MSc in CS, Oxford 2010. </LI> </UL> <UL> <LI>Jiewen Huang: <EM>Design and Implementation of the SPROUT Query Engine for Probabilistic Databases</EM> [<A href="/people/dan.olteanu/theses/Jiewen.Huang.pdf">PDF</A>]<BR> MSc by Research, Oxford 2009. </LI> </UL> <UL> <LI>Rasmus Wissmann: <EM>Tractable Queries with Inequalities on Probabilistic Databases</EM> [<A href="/people/dan.olteanu/theses/Rasmus.Wissmann.pdf">PDF</A>]<BR> MSc in CS, Oxford 2009. </LI> </UL> <UL> <LI><A title="data-collection" name="data-collection"></A>Smitha Mysore-Shankar: <EM>Learning Probabilistic Databases</EM><BR> MSc in CS, Oxford 2009.<BR> Data collection: <A href="/people/dan.olteanu/theses/Smitha.Shankar.urls.zip">URLs</A> to web-pages used to extract data and populate probabilistic databases </LI> </UL> <UL> <LI>Jiewen Huang: <EM>Efficient Query Evaluation for Tuple-Independent Probabilistic Databases</EM><BR> MSc in CS, Oxford 2008. </LI> </UL> <H3>Publications</H3> <UL> <LI><STRONG>Aggregates in Probabilistic Databases via Knowledge Compilation</STRONG>. [<A href="/people/Dan.Olteanu/papers/fho-vldb12.pdf">PDF</A>, <A href="/people/Dan.Olteanu/tutorials/vldb12-talk.pdf">slides</A>]<BR> Robert Fink, Larisa Han, and Dan Olteanu. <BR> In Very Large Data Bases (<SPAN>PVLDB), 5(5):490-501</SPAN>, Istanbul, 2012. </LI> </UL> <UL> <LI><STRONG>Ranking in Probabilistic Databases: Complexity and Efficient Algorithms</STRONG>. [<A href="/people/Dan.Olteanu/papers/ow-icde12.pdf">PDF</A>] <BR> Dan Olteanu and Hongkai Wen. <BR> In IEEE Int Conf on Data Engineering (ICDE), Washington DC, 2012. </LI> </UL> <UL> <LI> <STRONG> <A href="http://www.morganclaypool.com/doi/abs/10.2200/S00362ED1V01Y201105DTM016"> Probablistic Databases</A> </STRONG> <BR> Synthesis Lectures on Data Management, Morgan & Claypool, 2011, 180 pages. <BR> Dan Suciu, Dan Olteanu, Christopher Re, Christoph Koch. <BR> DOI: 10.2200/S00362ED1V01Y201105DTM016. </LI> </UL> <UL> <LI><STRONG>The Next Big Challenge in Data Management: Probabilistic Databases</STRONG>. <BR> Robert Fink and Dan Olteanu. <BR> Invited article in <A href="http://www.thinkdigit.com/">The Digit Magazine</A>, June 2011 (10th Anniversary Special), in print. <BR> More about the Digit Magazine in <A href="http://en.wikipedia.org/wiki/Digit_(magazine)">Wikipedia</A>. </LI> </UL> <UL> <LI><STRONG>Managing Probabilistic Data with SPROUT</STRONG>. <BR> Dan Olteanu. <BR> Article in the 2nd edition of Inspired Research published by the Department of Computer Science at the University of Oxford. June 2011. </LI> </UL> <UL> <LI><STRONG>SPROUT^2: A Squared Query Engine for Uncertain Web Data</STRONG> (Demonstration). [<A href="/people/dan.olteanu/papers/for-sigmod11.pdf">pdf</A>, <A href="/people/dan.olteanu/papers/for-sigmod11-poster.pdf">poster</A>, <A href="/people/dan.olteanu/sprout2/">web page</A> ] <BR> Robert Fink, Andrew Hogue, Dan Olteanu, and Swaroop Rath. <BR> In ACM SIGMOD, Athens, June 2011. </LI> </UL> <UL> <LI><STRONG>Providing Support for Full Relational Algebra in Probabilistic Databases</STRONG>. [<A href="/people/dan.olteanu/papers/for-icde11.pdf">pdf</A>]<BR> Robert Fink, Dan Olteanu, and Swaroop Rath. <BR> In IEEE Int Conf on Data Engineering (ICDE), Hannover, 2011. </LI> </UL> <UL> <LI><STRONG>On the Optimal Approximation of Queries using Tractable Propositional Theories</STRONG>. [<A href="/people/dan.olteanu/papers/fo-icdt11.pdf">pdf</A>, <A href="/people/Dan.Olteanu/tutorials/icdt11-talk.pdf">slides</A>]<BR> Robert Fink and Dan Olteanu. <BR> In Int Conf on Database Theory (ICDT), Uppsala, 2011. </LI> </UL> <UL> <LI><STRONG>Bridging the Gap Between Intensional and Extensional Query Evaluation in Probabilistic Databases</STRONG>.[<A href="/people/dan.olteanu/papers/jos-edbt10.pdf">pdf</A>] <BR> Abhay Jha, Dan Olteanu, and Dan Suciu.<BR> In Int Conf on Extending Data Base Technology (EDBT), Lausanne, March 2010. <BR><BR><EM> We extend the notion of tractable queries on (arbitrarily correlated) probabilistic databases to tractable data-query instances, where general hard queries can become tractable on restricted data instances. The query evaluation is separated into two steps: (1) the (possibly large) tractable data-query instance is evaluated first by efficient database techniques, and then (2) the (usually small) intractable residue is processed using inference techniques based on treewidth. </EM></LI> </UL> <UL> <LI><STRONG>Approximate Confidence Computation in Probabilistic Databases</STRONG>. [<A href="/people/dan.olteanu/papers/ohk-icde10.pdf">pdf</A>]<BR> Dan Olteanu, Jiewen Huang, and Christoph Koch. <BR> In IEEE Int Conf on Data Engineering (ICDE), Long Beach, March 2010. <BR><BR><EM>We enhance our query engine <A href="/projects/SPROUT/">SPROUT</A> with a deterministic approximation algorithm with error guarantees for confidence computation in (arbitrarily correlated) probabilistic databases, which is based on incremental compilation of query lineage into so-called decomposition trees that support linear-time probability computation. The compilation is incremental and we show experimentally that it can achieve a given approximation within a few steps. In case of tractable (hierarchial or inequality) queries, it always finishes in polynomial time. </EM></LI> </UL> <UL> <LI><STRONG>On Tractability of Inequality Queries over Probabilistic Databases and Counting Vertex Covers</STRONG>. <BR> Dan Olteanu and Rasmus Wissmann. <BR> technical report, September 2009. </LI> </UL> <UL> <LI><STRONG>Secondary-Storage Confidence Computation for Conjunctive Queries with Inequalities</STRONG>. [<A href="/people/dan.olteanu/papers/oh-sigmod09.pdf">pdf</A>] <BR> Dan Olteanu, Jiewen Huang. <BR> In ACM SIGMOD, Providence, 2009. <BR><BR><EM>This paper is the first to discuss tractability of conjunctive queries with inequalities in probabilistic databases. It presents a class of hard queries, and gives a scalable algorithm for tractable inequality queries that is based on ordered binary decision diagrams and is implemented in our query engine <A href="/projects/SPROUT/">SPROUT</A>. Our implementation passed the SIGMOD'09 repeatability and workability evaluation (RWE). Out of 63 accepted papers, 19 participated in RWE, and 10 passed RWE.</EM><BR><BR> Resources: <UL> <LI><A href="/people/dan.olteanu/code/oh-sigmod09-experiments.tgz">Scripts</A> for running the experiments. </LI> <LI><A href="/people/dan.olteanu/code/oh-sigmod09-sprout.tgz">Code</A> of the SPROUT query engine used in the experiments. </LI> </UL> </LI> </UL> <UL> <LI><STRONG>SPROUT: Lazy vs. Eager Query Plans for Tuple-Independent Probabilistic Databases</STRONG>. [<A href="/people/dan.olteanu/papers/ohk-icde09.pdf">pdf</A>] <BR> Dan Olteanu, Jiewen Huang, Christoph Koch. <BR> In IEEE Int Conf on Data Engineering (ICDE), Shanghai, April 2009. <BR><BR><EM> We show that the tractable hierarchical queries on probabilistic databases admit relational query plans with unrestricted join ordering, thereby overcoming the limitations of the safe plans of earlier work. In this framework, safe plans can be seen as one specific type of eager plans (where confidence computation is done as soon as possible). We extend relational plans with a new aggregation operator that is optimized using query signature and schema information (keys) to minimize the number of scans it needs to compute tuple confidences. This operator basically factorizes the query lineage into read-once functions (called 1OF formulas in the paper) in polynomial time. An extensive study on TPC-H queries is available <A href="papers/icde09queries.html">here</A>.</EM><BR><BR> Resources: <UL> <LI><A href="/people/dan.olteanu/papers/icde09queries.html">Case study</A>: Tractable TPC-H queries on probabilistic databases. </LI> <LI><A href="/people/dan.olteanu/tutorials/sprout.pdf">One-hour talk on the results of this paper</A>. </LI> <LI><A href="/people/dan.olteanu/code/ohk-icde10-sprout.tgz">Current version</A> of the SPROUT query engine. </LI> </UL> </LI> </UL> <UL> <LI><STRONG>Using OBDDs for Efficient Query Evaluation on Probabilistic Databases</STRONG>. [<A href="/people/dan.olteanu/papers/oh-sum08.pdf">pdf</A>] <BR> Dan Olteanu, Jiewen Huang. <BR> In <A href="http://www.kr.tuwien.ac.at/staff/lukasiew/sum08/">2nd Int. Conf. on Scalable Uncertainty Management (SUM)</A>, Napoli, Oct. 2008. Also in Proc. of Dagstuhl <A href="http://kathrin.dagstuhl.de/08421/Materials/">seminar #08421 on uncertainty management</A>, October 2008. <BR><BR><EM> This paper is the first to consider OBDDs for efficient evaluation of queries in probabilistic databases. It shows that for tractable hierarchical queries, the query lineage can be efficiently brought into one-occurrence form (that is, into an equivalent read-once function), where each variable occurs exactly once. For hard queries, existing results that bound the OBDD size in the pathwidth of the lineage apply immediately. </EM></LI> </UL> <UL> <LI><STRONG>Conditioning Probabilistic Databases</STRONG>. [<A href="http://arxiv.org/abs/0803.2212">pdf</A>] <BR> Christoph Koch, Dan Olteanu. <BR> In Very Large Data Bases (PVLDB), volume 1, 2008. <BR> Also ACM CORR Report cs.DB/0803.2212. <BR><BR><EM>This paper is the first to consider the problem of conditioning a probabilistic database outside of the context of graphical models. The core contribution is an exact confidence computation algorithm and its extension to achieve conditioning.</EM><BR><BR> Resources: <UL> <LI><A href="/people/dan.olteanu/code/confidence-computation.tgz">CODE:</A> Exact and approximate probability computation using world-set trees (now part of MayBMS). </LI> <LI><A href="/people/dan.olteanu/tutorials/vldb08-talk.pdf">Talk</A> given at <A href="https://www.cs.auckland.ac.nz/research/conferences/vldb08/index.php/VLDB_08">34th Int. Conf. on Very Large Data Bases (VLDB)</A>, Auckland, Aug 2008. </LI> </UL> </LI> </UL> <H3>Current team</H3> <UL> <LI>Robert Fink</LI> <LI>Dan Olteanu</LI> </UL> <P>Collaborators: Andrew Hogue (Google), Abhay Jha (U. of Washington), Christoph Koch (Cornell University, now at EPFL), Dan Suciu ((U. of Washington)</P> <P>Alumni: Larisa Han (MSc student), Jiewen Huang (research student), Andrei Marin (MSc student), Smitha Mysore-Shankar (MSc student), Lampros Papageorgiou (MSc student), Swaroop Rath (MSc student), Quentin Spencer-Harper (BSc student), Yunli Sung (MSc student), Rasmus Wissmann (MSc student), Hongkai Wen (research student)</P> <H3>Acknowledgments</H3> <P>From May 2009 to Oct 2012, SPROUT is partially supported by the Future and Emerging Technologies (FET) programme within the Seventh Framework Programme for Research of the European Commission, under the FETOpen grant agreement FOX, number FP7-ICT-233599. During the academic year 2008-2009, Jiewen Huang was generously supported by a scholarship from Cornell University. Starting April 2012, SPROUT is partially supported by the EPSRC grant PROQAW.</P> Department of Computer Science, University of Oxford Projects All oucl-All-SPROUT Fri, 10 Feb 2012 02:23:47 GMT http://www.cs.ox.ac.uk/projects/predictablesoftware/ The increasing complexity, scale and pervasive nature of software systems currently being deployed pose new challenges for software engineering. Traditional design and validation technologies are unable to deal with the needs for variability of context, for adaptability to changing scenarios, multiplicity of infrastructures and devices, and for addressing real-time and mobility issues. Software providers must have means to ensure and assess their confidence in the software system in advance of deployment. Consequently, techniques and tools able to predict and systematically validate the behaviour of large-scale networked software systems are necessary. Such tools, notably model checkers, have been developed within the formal methods community, and valuable lessons can also be learned from advances in theoretical computer science, and indeed in the applied mathematical analysis of complex systems in general. However, these cannot handle software adaptiveness, mobility, and system-of-systems issues. Predictable Software Systems is a paradigm which embodies the desire that pervasive complex software can be built from analysed heterogeneous components, can evolve and adapt over time, and yet where the resultant system would have predictable behaviour. The Predictable Software Systems (PSS) project is part of the Large-Scale Complex IT Systems research programme (LSCITS, pronounced else-its), a collaboration involving five institutions (www.lscits.org). The main focus of the LSCITS programme is on novel approaches and techniques for managing change. Department of Computer Science, University of Oxford Projects All oucl-All-predictablesoftware Fri, 10 Feb 2012 02:23:47 GMT