The success of Web 2.0 and CGM is based on tapping into the social nature of human interactions, by making it possible for people to voice their opinion, become part of a virtual community and collaborate remotely. If we take micro-blogging as an example, the growth in Twitter visits between 2008 and 2009 was over 1,000% and it is projected that by 2010 around 10% of all internet users will be on Twitter. This unprecedented rise in the volume and importance of online content has resulted in companies and individuals spending ever increasing amounts of time trying to keep up with relevant CGM. It is estimated that 700 person hours per year is the absolute minimum that companies and public services need to spend on CGM monitoring, online user engagement, and discovery of new information. This fellowship is about helping people to cope with the resulting information overload, through automatic methods that are capable of adapting to individual's information seeking goals and summarising briefly the relevant media and thus supporting information interpretation and decision making. Automatic text summarisation is key to our goal and consists of compressing the meaning of text documents while preserving the relevant information contained within them. While there has been a lot of research on well-authored texts such as news, summarisation of social media is still in its infancy, with research focused on product reviews. A key experimental finding has been that due to the characteristics of social media (product reviews in particular) it is better first to abstract the relevant information from the different documents and sites and then to use natural language generation to create a fluent text based on this information.In this fellowship I will investigate and evaluate new machine learning methods for personalised, abstractive multi-document summarisation across different social media. For example, diachronic summaries that combine Twitter posts, blog articles, and Facebook wall messages on a given topic. In contrast to previous work, we will pursue an inter-disciplinary approach, which will help us study the social dimension of CGM summarisation and establish actual user needs. The second research challenge is that the algorithms need to be robust in the face of this noisy, jargon-full and dynamic content, as well as needing models capable of representing the contradictory and strongly temporal nature of CGM. A key novel contribution of our work is personalising the summaries, based on a model of user interests, goals, and social context. Issues such as trustworthiness, privacy, and online communities (with their hubs and authorities) will also play an important role. The fourth research challenge is to generate personalised abstractive summaries that can help users with sensemaking and content interpretation. An exciting element of my research will be in studying the different kinds of summaries that are useful for a variety of real users (companies, journalists, and the general public) through multi-disciplinary collaborations with the Press Association, British Telecom, the Oxford Internet Institute, and Sheffield's Department of Journalism. A key project deliverable will be a publicly available browser plugin that provides easy access to the automatically generated summaries. This will allow me to evaluate the project results with real users, on a large scale. It will also provide a new evaluation challenge for the Natural Language Generation community, as researchers will be able to compare their summarisers against those delivered by our open-source algorithms. Last but not least, the fellowship covers not only foundational multi-disciplinary research but it also tests the results in several Digital Economy pilot experiments involving commercial partners (The Press Association, British Telecom, Fizzback).

This Fellowship will enable the rigorous study of a fragile structural form that has long left the comfortable confines of the laboratory scale and is increasingly critical to our renewable energy independence. This Fellowship will, for the first ever time: develop open-source solvers for the high-performance simulation of structural systems with sharply nonlinear behaviour suffering from numerical deterioration in partnership with the Edinburgh Parallel Computing Centre (EPCC); develop protocols for the digital twinning of massive shell structures where the quality of the twinned midsurface is paramount and sub-mm geometric features can be critical, in partnership with reality capture specialists Leica Geosystems (LGS); gather the first terabyte-sized datasets of digital twin inputs representing state-of-the-art offshore wind support structures based on unprecedented access to facilities planned for construction starting in 2025 granted by project partners Siemens Gamesa Renewable Energy (SGRE), ScottishPower Renewables (SPR) and COWI; complete the scientific understanding of the nonlinear response of very long tubular structural forms prone to ovalisation phenomena; generate extensive datasets of synthetic buckling resistances of digitally-twinned shells; calibrate actual safety margins of current and future planned offshore wind support structures and disseminate this within the international Eurocode design framework; ('Plus') found a permanent indexed data journal to accumulate empirical and numerical dataset pairs for the wider computational engineering community to validate simulations used in research and safety-critical design. The open-source software development will push the boundaries of computational structural engineering and support an emerging research culture increasingly employing digital twinning. The financial benefits of quantifying actual safety margins of current and future-scale offshore wind support structures are significant: a single modern tower saved from failure saves ~£2M, while even a ~10% reduction in steel saves ~£10M across a 100-tower offshore installation (assuming ~£1k / tonne for structural steel, not including carbon cost).

Data science and artificial intelligence will transform the way in which we live and work, creating new opportunities and challenges to which we must respond. Some of the greatest opportunities lie in the field of human health, where data science can help us to predict and diagnose disease, determine the effectiveness of existing treatments, and improve the quality and affordability of care. The Oxford EPSRC CDT in Health Data Science will provide training in: - core data science principles and techniques, drawing upon expertise in computer science, statistics, and engineering - the interpretation and analysis of different kinds of health data, drawing upon expertise in genomics, imaging, and sensors - the methodology and practice of health data research, drawing upon expertise in population health, epidemiology, and research ethics The training will be provided by academics from five university departments, working together to provide a coordinated programme of collaborative learning, practical experience, and research supervision. The CDT will be based in the Oxford Big Data Institute (BDI), a hub for multi-disciplinary research at the heart of the University's medical campus. A large area on the lower ground floor of the BDI building will be allocated to the CDT. This area will be refurbished to provide study space for the students, and dedicated teaching space for classes, workshops, group exercises, and presentations. Oxford University Hospitals NHS Foundation Trust (OUH), one of the largest teaching hospitals in the UK, will provide access to real-world clinical and laboratory data for training and research purposes. OUH will provide also access to expertise in clinical informatics and data governance, from a practical NHS perspective. This will help students to develop a deep understanding of health data and the mechanisms of healthcare delivery. Industrial partners - healthcare technology and pharmaceutical companies - will contribute to the training in other ways: helping to develop research proposals; participating in data challenges and workshops; and offering placements and internships. This will help students to develop a deep understanding of how scientific research can be translated into business innovation and value. The Ethox Centre, also based within the BDI building, will provide training in research ethics at every stage of the programme, and the EPSRC ORBIT team will provide training in responsible research and innovation. Ethics and research responsibility are central to health data science, and the CDT will aim to play a leading role in developing and demonstrating ethical, responsible research practices. The CDT will work closely with national initiatives in data science and health data research, including the ATI and HDR UK. Through these initiatives, students will be able to interact with researchers from a wide network of collaborating organisations, including students from other CDTs. There will also be opportunities for student exchanges with international partners, including the Berlin Big Data Centre. Students graduating from the CDT will be able to understand and explore complex health datasets, helping others to ask questions of the data, and to interpret the results. They will be able to develop the new algorithms, methods, and tools that are required. They will be able to create explanatory and predictive models for disease, helping to inform treatment decisions and health policy. The emphasis upon 'team science' and multi-disciplinary working will help to ensure that our students have a lasting, positive impact beyond their own work, delivering value for the organisations that they join and for the whole health data science community.