Intelligent and Adaptive Systems Research Group Past Seminars and Events
4 July 2014
Speaker: Teemu H. Laine
Title: Ubilife – Research on smart education and wellbeing
23 January 2014
Speakers: Jane Sinclair and Claire Rocks
Title: Computing for Teachers MOOC
19 November 2013
Speakers: Reem Alfayez, Alaa A. Qaffas
Title: Research presentation
- Presentation by Reem Alfayez
- Presentation and paper by Alaa A. Qaffas
12 November 2013
Speaker: Maria Liakata
Title: A discourse-driven content model for summarising scientific articles evaluated in a complex question answering task
We present a method which exploits automatically generated scientific discourse annotations to create a content model for the summarisation of scientific articles. Full papers are first automatically annotated using the CoreSC scheme, which captures 11 content-based concepts such as Hypothesis, Result, Conclusion etc at the sentence level. A content model which follows the sequence of CoreSC categories observed in abstracts is used to provide the skeleton of the summary, making a distinction between dependent and independent categories. Summary creation is also guided by the distribution of CoreSC categories found in the full articles, in order to adequately represent the article content. Finally, we demonstrate the usefulness of the summaries by evaluating them in a complex question answering task. Results are very encouraging as summaries of papers from automatically obtained CoreSCs enable experts to answer 66% of complex content-related questions designed on the basis of paper abstracts. The questions were answered with a precision of 75%, where the upper bound for human summaries (abstracts) was 95%.
5 November 2013
Speaker: Richard Orme
Title: A new era of reading for people with impairments, powered by technology
The recent rapid development of digital publishing is offering unprecedented access to books for people with print disabilities, offering readers the ability to personalise the experience to meet their individual requirements. This engaging seminar will explore the technologies and collaborations that make this possible, complete with examples, case studies and, despite the many opportunities for disaster, live demonstrations.
Richard Orme is an international leader in the area of digital inclusion for people with disabilities, especially active in the publishing, broadcasting and technology sectors.
30 October 2013
Speakers: Mmaki Jantjies, Maram Al Zaidi, Monisa Abdul Wahab
Title: Paper & research presentation
- Presentation & paper by Mmaki Jantjies
- Presentation & notes by Maram Al Zaidi
- Presentation & paper by Monisa Abdul Wahab
22 October 2013
Speaker: Jane Sinclair
Title: Paper presentation & discussion
15 October 2013
Speakers: Karen Stepanyan, Suncica Hadzidedic, George Gkotsis
Title: Interesting papers presentation
- Presentation by Karen Stepanyan (paper)
- Presentation by Suncica Hadzidedic (paper)
- Presentation by George Gkotsis (paper)
8 October 2013
Speaker: IAS members
Title: Setup of IAS seminars for upcoming term
The first meeting was focused on discussing the different directions of the seminars. There were several ideas proposed by the attendants, listed below:
- Research methodology tutorial/presentation (e.g. invite someone from a workshop) (RSSP)
- IAS members may find useful the tutorials on how to present/write a paper offered by RSSP.
- Journal club
- People read the same paper and one presents
- People read the same paper and all of them discuss
- Student-focused. Make sure we get feedback from the senior academic members.
- Duration should drop to 20’. Have 10’ for Q&A and discussion.
- Combination of a 30’ presentation and open, brainstorm discussion
- Speed dating
- Assemble a list of relevant/interesting conferences or journals
- Wrap up what kind of content we have on our website
- Tutorial/presentation on tools and technologies used in everyday
- One hour having as many possible presenting interesting papers we read recently.
12 November 2012
Speaker: Dom Orchard, Computer Laboratory, Cambridge University
Title: Programs in Context
Abstract:
Computer programs are not executed in a vacuum, they are run in diverse environments where the execution context may change a program’s capabilities, behaviour, and results. Notions of context in programming are wide ranging, from user- and application-level notions such as a device’s GPS location or local resources, to low-level programming notions such as binding scopes or data-structure traversals. This talk presents a general program analysis for contextual properties of programs. This analysis can be used to provide information on program requirements, aid program correctness, and guide optimisation. The approach can be specialised to various notions of context: examples involving array traversals and distributed resources will be shown. Research on a general semantics for context-dependence will also be informally described.
5 November 2012
Speakers: Jane Prewett and Graham Hewitt, Research Support Services, University of Warwick
Title: Research Ethics
Abstract:
For any research, ethical considerations must be taken into account and approval gained beforehand via official institutional channels. For us in the IAS group, many research exercises will involve elements for which ethical consent will be required, even if it is only a student questionnaire or user evaluation of a system we have developed.
Procedures at Warwick have changed recently and are currently significantly different to the position a few months ago. Many IAS members may be affected by this. We are very pleased to welcome Jane and Graham for this seminar which will:
- explain the importance of ethical approval for research;
- explain the university's new procedures
- talk about ammendments to these which will also affect us
- provide an opportunity for questions and discussion
29 October 2012
Speaker: Dr Craig Stewart, Department of Computer Science, University of Nottingham
Title: Evaluating a Cultural Stereotype
22 October 2012
Speaker: Tim Monks, Department of Computer Science, University of Warwick
Title: Supporting Empirical Modelling in the Browser
Abstract:
This talk will cover JsEden, a tool developed to support interactive model building in web browsers. The tool was conceived and implemented by the speaker as an MSc dissertation project in 2011, but has already been developed further by students and researchers at the University of Warwick. The talk will briefly cover the agenda of Empirical Modelling, how JsEden supports this agenda and fits into the current family of similar tools, details about the implementation of JsEden and lastly an overview of the plans for further development.
15 October 2012
Speaker: Dr Meurig Beynon, Department of Computer Science, University of Warwick
Title: Realising Software Development as a Lived Experience
Abstract:
This essay discusses software development from the perspective of Empirical Modelling (EM), an approach to computing that draws on the construals of David Gooding, Bruno Latour’s vexing notion of construction and William James’s radical empiricism. It argues that effective software development must embrace semantic principles radically different from those endorsed by the traditional perspective on software that is based on computational thinking. Of paramount importance is the immediacy of the developer’s experience of the relationship between software as an artefact on the computer and software as an agency in the world.
6 July 2012
Speaker: Prof. Dave West, New Mexico Highlands University School of Business
Title: Design Thinking and Empirical Modelling: a potential synergy?
Abstract:
The challenges posed for software developers by ultra-large scale and complex adaptive systems are not amenable to solutions using traditional computer science and software engineering approaches. Problems in this new domain are characterized by vague and ambiguous requirements, unknown variables and relationships, and satisfaction constraints that are more subjective than objective. Designers (architects and applied artists like graphic and industrial designers) seem to have developed approaches and methods for solving this kind of problem. In this short talk I will briefly describe essential elements of this approach and what they offer to software development and then see if I can establish an initial mapping to some of the essential aspects (as I understand them) of EmpiricalModelling to see if there is any basis for exploration and research.
David West, Ph.D., is a leading educator and trainer in object-oriented software development with extensive experience consulting for Fortune 500 companies. He is a professor at New Mexico Highlands University School of Business, where he is currently teaching an object-based curriculum and developing an innovative software development apprenticeship program based on object and agile principles. Dr. West also helped establish a software development track for computer science graduate students at the University of New Mexico. He founded and directed the Object Lab at the University of St. Thomas and cofounded the original Object Technology User Group, with more than 500 members.
18 June 2012
Speaker: Dr Jason Nurse, Department of Computer Science, University of Oxford
Title: Using Information Trustworthiness Advice in Decision-Making
Abstract:
In a society at the brink of information overload, using a measurement of trustworthiness to focus attention and ultimately reduce risks faced by individuals is an increasingly attractive option in supporting well-conceived decisions. As such, this research seeks to advance discussions on trustworthiness and decision-making research by critically investigating individual's ability to cognitively combine trustworthiness measures and the information content that they relate to, to make decisions. This is an often assumed reality but one that is lacking focused analysis in the socio-technical field. In our experiments, as we present trustworthiness information using visualisations on a computer screen, we also conduct a secondary assessment of a range of visualisation techniques to determine whether there are any better or generally preferred approaches to support decisions. Findings from both evaluations are relatively positive and insightful, and amongst other aspects, reaffirm humans as optimal assessors and identify a particularly strong dependence on trustworthiness levels in influencing to decision-making. This talk is based on a paper that will be presented at the 2012 Workshop on Socio-Technical Aspects in Security and Trust. Broadly, this is research is part of the EPSRC and TSB-funded project, TEASE (Trust Enabling Augmented-Reality Support for Information-Environments).
14 May 2012
Speaker: Lei Shi, Department of Computer Science, University of Warwick
Title: Social Networking and Adaptation
30 April 2012
Speaker: Dr Soha Maad, University College Cork
Title: Leveraging ICT Background in Healthcare Research and Practice
Abstract:
Soha Maad did her PhD in Computer Science with the Empirical Modelling group at the University of Warwick. Her talk will briefly overview her activities since she left Warwick University in 2002 and will focus on how she leveraged her ICT background in penetrating the healthcare sector and addressing key challenges identified in various R&D agendas. Focusing on primary care, Soha Maad will present her research on developing a global platform for primary care supported by the Royal College of Surgeons in Ireland. Leveraging ICT to promote active and healthy ageing among various target groups including the disabled is a major research interest for Soha Maad. Soha Maad collaborated with various prominent research groups in healthcare (including Phillips healthcare, Technical University Eindhoven, and Synergetics) in preparing research proposals in the area of telemedicine deployment and service innovation for active and healthy ageing. Her future vision is to build entrepreneurial skills and develop spin off companies targeting a global healthcare market.
12 March 2012
Speaker: Anna Jordanous, Kings College London
Title: Evaluating Computational Creativity: A Standardised Procedure for Evaluating Creative Systems and its Application
Abstract:
No methodology has been accepted as standard for evaluating the creativity of a system in the field of computational creativity and the multi-faceted and subjective nature of creativity generates substantial definitional issues. Evaluative practice has developed a general lack of rigour and systematicity, hindering research progress.
To tackle these issues, the SPECS methodology is proposed: a Standardised Procedure for Evaluating Creative Systems. SPECS is a standardised and systematic methodology for evaluating computational creativity. It is flexible enough to be applied to a variety of different types of creative system and adaptable to specific demands in different types of creativity. In the three-stage process of evaluation, researchers are required to be specific about what creativity entails in the domain they work in and what standards they test a system's creativity by. To assist researchers, definitional issues are investigated and a set of components representing aspects of creativity is presented, which was empirically derived using computational linguistics analysis. These components are offered as a general definition of creativity that can be customised to account for any specific priorities for creativity in a given domain.
SPECS is applied in a case study for detailed comparisons of the creativity of four musical improvisation systems, identifying which systems are more creative than others and why. In a second case study, SPECS is used to capture initial impressions on the creativity of systems presented at a 2011 computational creativity research event.
Five systems performing different creative tasks are compared and contrasted. These case studies exemplify the valuable information that can be obtained on a system's strengths and weaknesses. SPECS gives researchers vital feedback for improving their systems' creativity, informing further progress in computational creativity research.
5 March 2012
Speaker: Saba Khalil Toor, Department of Computer Science, University of Warwick
Title: Erroneous Bibliographies in Students’ Course Work: Is There A Solution?
Abstract:
This research work is aimed to investigate reasons for the presence of erroneous bibliographies and reference list with in students’ course work, and to propose a viable solution. In other word we are trying to determine why students of higher education are presenting reference lists and bibliographies that have incorrect or ambiguous information. This is a critical factor for any scholarly work and error in these lists can cause retrieval of such information either impossible or very difficult.
To investigate this issue an in-depth research is being conducted. Both quantitative and qualitative research methods are being employed i.e. a hybrid method of research is being used to acquire thorough understanding of this aspect. Scope of this research is limited to the academia, and we are targeting students and teachers of undergraduate and post graduate levels. Thus the two stake holders are included in this investigation.
13 February 2012
Speaker: Dr George Gkotsis, Department of Computer Science, University of Warwick
Title: CoPe_it!: Argumentative Collaboration towards Learning
Abstract:
CoPe_it! is an innovative web-based tool that supports collaboration and augments learning among members of diverse communities. The tool fosters the means to manage individual and collective knowledge during a sense-making or a decision-making session. The presentation will focus on its applicability for Communities of Practice (CoPs) by examining it from both a learning and an argumentative collaboration perspective. Arguing that argumentation is an essential element of the learning process, related design issues will be discussed.
30 January 2012
Speaker: Dr Karen Stepanyan, Department of Computer Science, University of Warwick
Title: BlogForever: Aiming to Preserve the Value of Blogs
Abstract:
The BlogForever collaborative European research project completed its first deliverables. This presentation covers some of the initial results that inform the primary goals of the project in developing facilities for robust digital preservation, management and dissemination of blogs. Unlike other web preservation initiatives, such as Internet Archive, BlogForever aims to capture the dynamic and continuously evolving nature of weblogs, their network and social structure. This presentation covers some of the challenges and potential solutions that are relevant to blog preservation initiatives with a wide range of stakeholders. Given the fact that lossless preservation of blogs in their entirety remains unrealistic and unsustainable this presentation is aiming to generate a discussion on what to preserve and why.
23 January 2012
Speaker: Dr Maurice Hendrix, Serious Games Institute
Title: Medical Education Content Sharing and Repurposing
Abstract:
A large amount of medical educational content (e.g. tests, videos, simulations etc.) is available within academic institutions and on the web. However all this information is not automatically finding its way into e-learning systems. The mEducator project is a best practice network, set up to examine existing standards and solutions and to review how the situation can be improved. The lack of standardized content sharing mechanisms has been identified as one of the main issues. mEducator 3.0, based on the Open Linked Education philosophy, has been implemented using a set of Semantic Web Services and a number of different interfaces to it. We will examine an evaluation of mEducator3.0 in terms of its performance, usability and viability.
Once educational content is shared effectively, an important issue is their re-use. As simulations and serious games are often difficult to re-purpose (re-use for a different purpose than the original author had), we will present mEitor, a tool that has been developed to facilitate the re-purposing of serious games, with a minimum of technical knowledge.
5 December 2011
Speaker: Dr Meurig Beynon, Department of Computer Science, University of Warwick
Title: A Future for EM
28 November 2011
Speaker: Dr Steve Russ, Department of Computer Science, University of Warwick
Title: Knowing and Computing
14 November 2011
Speaker: Jerry Foss, Birmingham City University
Title: Personalized Networked Video
7 November 2011
Speaker: Joshua Scotton, Department of Computer Science, University of Warwick
Title: Adaptation Types within Adaptive Educational Hypermedia
31 October 2011
Speaker: Russell Boyatt, Department of Computer Science, University of Warwick
Title: Introducing e-Prints
Abstract:
Something we should all know about - both in terms of the practicalities and its importance as a dissemination tool. A chance for old hands and new members to find out more.
24 October 2011
Speaker: Shanshan Yang, Department of Computer Science, University of Warwick
Title: An Effective Services Framework for Sharing Educational Resources
31 May 2011
Speaker: Mmaki Jantjies, Department of Computer Science, University of Warwick
Title: Bilingual Mobile Learning in South African Schools
Abstract:
Mobile learning has been seen as a tool that can be used to mitigate the digital divides in rural South African schools. From basic mobile phones to complex smart phones, there are various platforms and techniques of using this technology to support education. The method in which the concepts are developed often determines the success of its intended purpose. This seminar will examine ways of contextualising mobile learning in a diversely linguistic country with the aim of enhancing existing learning material.
24 May 2011
Speaker: Dr Karen Stepanyan, Department of Computer Science, University of Warwick
Title: BlogForever: The Road Ahead
Abstract:
BlogForever is a collaborative research project funded by the European Union as part of the Seventh Framework Programme (FP7). The main goal of the project is to develop facilities for robust digital preservation, management and dissemination of weblogs. The development of these facilities will enable capturing the dynamic and continuously evolving nature of weblogs, their network and social structure, and the exchange of concepts and ideas they foster.
Among the leading beneficiaries of BlogForever is the University of Warwick. The initial and the primary role of the university is to lay foundations by exploring weblog structure and semantics and proposing a generic weblog data model. As a leading partner, the university will contribute to the following stages of the project for defining a digital preservation policy for weblogs, developing and piloting a digital repository of weblogs. In addition to providing an overview of the project, this presentation is intended to seek feedback and suggestions for the road ahead.
17 May 2011
Speaker: Prof. John Traxler, Wolverhampton University
Title: Africa, Learning and Mobile Technology: The Elephant in the Room
Abstract:
There is much discussion and much interest in the capacity of mobile technologies to deliver, support and enhance learning for the disenfranchised, the disadvantaged and the developing communities and regions of the world especially in Africa.
Much of this discussion, interest and activity is however uncritical, simplistic and poorly synthesised.
In general the argument for using these technologies to address educational disadvantage is plausible, self-evident and straightforward: their ownership and acceptance are near-universal and cut across most notions of digital divides; their use is based around robust sustainable business models; they are, unlike other ICTs, found at the bottom-of-the-pyramid, amongst the next billion subscribers; they deliver information, ideas and, increasingly, images. And there are no other options!
There is furthermore a rapidly increasing ownership of more powerful handsets in the developing world, decreasing real costs of this hardware and connectivity, increasing coverage of higher specification networks in these regions and renewed activity of donors and of corporates representing publishing, handsets, services and infrastructure looking for business models based on the educational use of mobile devices.
These various communities, including computer scientists and researchers, necessary actors in facilitating successful learning using mobile devices and technologies, each come with considerable potential but often inappropriate contributions, partial understandings and flawed assumptions. This seminar will explore the extent to which their optimism is misplaced and where computer science might fit into the picture.
18 March 2011
Speakers: William Dangerfield and Matthew Carter, Department of Computer Science, University of Warwick
Title: Some More Student Experiences of Empirical Modelling
Abstract:
This seminar is a follow-up to last week's seminar, and will have the same overall format (but omitting the introduction and leaving more room for discussion). It showcases the work of another two students who attended the CS405 "Introduction to Empirical Modelling" module this year. Over the last seven years, the coursework for CS405 has consisted of a submission for the online Warwick Electronic Bulletin on Empirical Modelling ("WEB-EM"). Each submission comprises a short conference-style presentation and an associated modelling study based on a theme chosen by the student. The submissions are especially useful as an indication of how far the principles and tools of EM can be exploited by students who join the module in October with no prior knowledge, and have to demonstrate hat they have learnt early in Term 2. In this seminar, William and Matthew will present modelling studies with potential for practical application and comment on their experience of EM. Construals based both on using the established tkeden interpreter and the more recently developed Cadence interpreter will be featured. As in the previous seminar, Steve Russ will act in the role of a respondent, mediating the ideas behind the modelling studies to members of the IAS research group.
10 March 2011
Speakers: Seb Sidbury and David Evans, Department of Computer Science, University of Warwick
Title: Some Student Experiences of Empirical Modelling
Abstract:
Over the last seven years, the coursework for the "Introduction to Empirical Modelling" module CS405 has consisted of a submission for the online Warwick Electronic Bulletin on Empirical Modelling ("WEB-EM"). Each submission comprises a short conference-style presentation and an associated modelling study based on a theme chosen by the student. The submissions are especially useful as an indication of how far the principles and tools of EM can be exploited by students who join the module in October with no prior knowledge, and have to demonstrate that they have learnt early in Term 2. In this seminar, two CS405 students will present their modelling studies and comment on their experience of EM. Construals based both on using the established tkeden interpreter and the more recently developed Cadence interpreter will be featured. Steve Russ will act in the role of a respondent, mediating the ideas behind the modelling studies to members of the IAS research group.
3 March 2011
Speaker: Dr. Amirah Ismail, Universiti Kebangsaan Malaysia
Title: A Metametadata Framework to Support Semantic Searching of Pedagogic Data
Abstract:
A brief introduction to our Faculty of Information Science and Technology, University Kebangsaan Malaysia as well as our current research centres and research groups. The main focus of our research under the management of the Centre For Artificial Intelligence Technology (CAIT) are Artificial Intelligence, Data Mining, Pattern Recognition and Knowledge Technology. The center’s main purpose is the application of Artificial Intelligence techniques and methods to support various disciplines through research, teaching and consultancy. Our research is funded in part by grants (both internally (UKM) and externally (GUP, FRGS, ScienceFUND, IRPA)) and contract researches as well as local industry and other research institute. Our research group, the Knowledge Technology (KT) research group focused on research relating to machine processable semantics and the used of semantic technology in various advance applications. The domain these groups are currently working on is in digital library, e-learning and health informatics. Among the research conducted by this research group are semantic searches, ontology construction from text, semantic digital library, ontology based health information system and statictical Malay parser. My Phd Thesis entitled A Metametadata Framework to Support Semantic Searching of Pedagogic Data focuses on a novel method for semantic searching and retrieval of information about learning materials. Metametadata encapsulate metadata instances by using the properties and attributes provided by ontologies rather than describing learning objects. A novel metametadata taxonomy has been developed which provides the basis for a semantic search engine to extract, match and map queries to retrieve relevant results. The use of ontological views is a foundation for viewing the pedagogical content of metadata extracted from learning objects by using the pedagogical attributes from the metametadata taxonomy. My PhD research has literally support part of our current research group, KT research group as well as our centre’s aim and motivating our post grads to explore more into the semantic searches and algorithms within the e-learning environment.
11 February 2011
Speaker: Dr. Maurice Hendrix
Title: Supporting Authoring of Adaptive Hypermedia
Abstract:
It is well-known that students benefit from personalised attention. However, requently teachers are unable to provide this, most often due to time constraints. An Adaptive Hypermedia (AH) system can offer a richer learning experience, by giving personalised attention to students. The authoring process, however, is time consuming and cumbersome. Our research explores the two main aspects to authoring of AH: authoring of content and adaptive behaviour. The research proposes possible solutions, to overcome the hurdles towards acceptance of AH in education.
Automation methods can help authors, for example, teachers could create linear lessons and our prototype can add content alternatives for adaptation.
Creating adaptive behaviour is more complex. Rule-based systems, XML-based conditional inclusion, Semantic Web reasoning and reusable, portable scripting in a programming language have been proposed. These methods all require specialised knowledge. Hence authoring of adaptive behaviour is difficult and teachers cannot be expected to create such strategies. We investigate three ways to address this issue.
- Reusability: We investigate limitations regarding adaptation engines, which influence the authoring and reuse of adaptation strategies. We propose a meta-language, as a supplement to the existing LAG adaptation language, showing how it can overcome such limitations.
- Standardisation: There are no widely accepted standards for AH. The IMS-Learning Design (IMS-LD) specification has similar goals to Adaptive Educational Hypermedia (AEH). Investigation shows that IMS-LD is more limited in terms of adaptive behaviour, but the authoring process focuses more on learning sequences and outcomes.
- Visualisation: Another way is to simplify the authoring process of strategies using a visual tool. We define a reference model and a tool, the Conceptual Adaptation Model (CAM) and GRAPPLE Authoring Tool (GAT), which allows specification of an adaptive course in a graphical way. A key feature is the separation between content, strategy and adaptive course, which increases reusability compared to approaches that combine all factors in one model.
10 February 2011
Speaker: Joshua Scotton, Department of Computer Science, University of Warwick
Title: The Adaptive Display Environment for Adaptive Hypermedia
Abstract:
A great body of previous research shows that personalization is a highly desirable method for online content delivery, based on the offline model of one to one personal interactions. Many different engines have been proposed to support such personalization on the Web. They all however have some weaknesses, which may have been partially responsible for the slow uptake of these promising techniques. Therefore, this paper discusses the advantages and problems with current adaptation engines, as well as proposes a new adaptation engine, ADE, the Adaptive Display Environment for Adaptive Hypermedia, and compares it with arguably the most famous and flexible current adaptation engine delivered by research, AHA!. This comparison is done from both a technical point of view, as well as with the help of a user study, and finds that ADE is significantly better received than AHA! in a user’s overall personal preference, as well as ADE’s layout and navigational interfaces.
27 January 2011 (Departmental Seminar)
Speaker: Prof. Wolfgang Nejdl, Hannover University
Title: Web of People -- Improving Search on the Web
Abstract:
More and more information is available on the Web, and the current search engines do a great job to make it accessible. Yet, optimizing for a large number of users, they usually provide good answers only to “most of us", and have yet to provide satisfying mechanisms to search for audiovisual content. In this talk I will present ongoing work at L3S addressing these challenges. I will start by giving a brief overview of Web Science areas covered at L3S, and the main challenges we adress in these areas, with the Web of People as one important focal point of our research, as well as Web Information Management and Web Search.
In the second part of the talk, I will discuss search for audiovisual content, and how to make this content more accessible. As many of our algorithms focus on exploiting user generated information, I will discuss what kinds of tags are used for different resources and how they can help for search. Collaborative tagging has become an increasingly popular means for sharing and organizing Web resources, leading to a huge amount of user generated metadata. These tags represent different aspects of the resources they describe and it is not obvious whether and how these tags or subsets of them can be used for search. I will present an in-depth study of tagging behavior for different kinds of resources - Web pages, music, and images. I will also discuss how to enrich existing tags through machine learning methods, to provide indexing more appropriate to user search behavior.
Biography:
Prof. Dr. Wolfgang Nejdl (born 1960) has been full professor of computer science at the University of Hannover since 1995. He received his M.Sc. (1984) and Ph.D. degree (1988) at the Technical University of Vienna, was assistant professor in Vienna from 1988 to 1992, and associate professor at the RWTH Aachen from 1992 to 1995. He worked as visiting researcher / professor at Xerox PARC, Stanford University, University of Illinois at Urbana-Champaign, EPFL Lausanne, and at PUC Rio. Prof. Nejdl heads the L3S Research Center as well as the Distributed Systems Institute / Knowledge Based Systems , and does research in the areas of search and information retrieval, information systems, semantic web technologies, peer-to-peer infrastructures, databases, technology-enhanced learning and artificial intelligence. Some recent projects in the L3S context include the PHAROS Integrated Project on audio-visual search, the OKKAM IP focusing on entities on the Web, the Digital Library EU project LiWA, coordinated by L3S, which investigates Web archive management and advanced search in such an archive, and the FET IP project LivingKnowledge, which is developing algorithms and methods to handle and exploit diversity, bias and opinion on the Web. Another new project, GLOCAL, focuses on event-based indexing of multimedia data on the web. Wolfgang Nejdl published more than 250 scientific articles, as listed at DBLP, and has been program chair, program committee and editorial board member of numerous international conferences and journals, most recently including the role of PC chair for WWW'09 in Madrid, PC chair for WSDM'11 in HongKong, and general chair for ICDE'11 in Hannover, see also http://www.kbs.uni-hannover.de/~nejdl/
21 January 2011
Speaker: Fawaz Ghali, Department of Computer Science, University of Warwick
Title: Social Personalized E-Learning Framework
Abstract:
This thesis discusses the topic of how to improve adaptive and personalized e-learning, in order to provide novel learning experiences. A recent literature review revealed that adaptive and personalized e-learning systems are not widely used. There is a lack of interoperability between adaptive systems and learning management systems, in addition to limited collaborative and social features. First of all, this thesis investigates the interoperability issue via two case studies. The first case study focuses on how to achieve the interoperability between adaptive systems and learning management systems using e-learning standards, and the second case study focuses on how to augment e-learning standards with adaptive features. Secondly, this thesis proposes a new social framework for personalized e-learning, in order to provide adaptive and personalized e-learning platforms with new social features. This is not just about creating learning content, but also about developing new ways of learning. For instance, in the presented vision, adaptive learning does not refer to individuals only, but also to groups. Furthermore, the boundaries between authors and learners become less distinct in the so-called Web 2.0 context. Finally, a new social personalized prototype is introduced based on the new social framework for personalized e-learning in order to test and evaluate this framework. The implementation and evaluation of the new system were carried through a number of case studies.
20 January 2011
Speaker: Jonny Foss, Department of Computer Science, University of Warwick
Title: The GRAPPLE FP7 Project Authoring Tool, GAT
Abstract:
The talk will describe the research ideas behind a novel approach to creating authoring tools for adaptive e-learning courses, created within the GRAPPLE FP7 STREP project. Specifically, the talk will show how a number of separate innovative components can be combined into a coherent authoring tool, the GRAPPLE Authoring Tool (GAT). GAT is now at its second iteration, based on some initial evaluations of the first version of the tool, half-way through the project. The tool will also be demonstrated, and participants will be encouraged to try it out during the event. For this purpose, participants will need to bring their own laptops, and ensure they have network access.
9 December 2010
Speaker: Dr. Maurice Hendrix, Wellcome Trust Sanger Institute
Title: Helpful Genome Browsing
Abstract:
The Ensemble website is a genome browser. It, makes complex genomics information available, and is capable of displaying a lot of this information in a variety of ways. Over the years, the website has become an important resource to the research community and continues to expand the data it provides and the various ways it displays this data. As an inevitable result of this ever growing amount of data, and number of displays, using the Ensembl website can be a challenge to novice users. The Ensembl web site therefore has a help system, which provides explanations about the data and displays. At present many users however still struggle to use the site and contact the outreach teams with questions. Many of these questions are answered in the help pages. It is clear that the current verbose one size fits all approach the information is presented in poses challenges to users who are lost on the website and refer to the help pages. A verbose one-size fits all help with a lot of navigational option may cause the user to become the lost in hyperspace in the help itself.
An additional challenge is the existence of sites based on the Ensembl site such as Vega and Ensembl Genomes. These websites are similar in function, but have a different target audience and display different data. Therefore the information on the help pages of the Ensembl websites also applies to these derived sites in many places. However there are often subtle, and sometimes not so subtle, differences. The approach of taking the Ensembl help and manually sifting through it for these subtle differences is tedious and time consuming.
2 December 2010
Speaker: Joshua Scotton, Department of Computer Science, University of Warwick
Title: Delivering Adaptive Content
Abstract:
The presentation will begin with an overview of adaptive hypermedia. This will introduce different types of adaptation and how adaptation languages can be used to specify adaptation behaviour within a hypermedia system. Different delivery systems for adaptive hypermedia and the development of the Adaptive Display Environment (ADE) will be described. A method for simplifying and increasing the reuse of adaptation specifications will also be discussed. Future research work, particularly that of extending the LAG adaptation language will be outlined in the final part of the presentation.
26 November 2010
Speaker: Eric Braude, Boston University, Metropolitan College
Title: Cumulative Subgoal Fulfillment: A New Approach to Developing Software
Abstract:
This talk will show how a few principles of physical construction, so self-evident that they are usually unremarked, can be very useful for software construction. This talk will show how the result, formulated as Cumulative Subgoal Fulfillment, applies to classic computer science examples, to Naur’s well-known but error-prone text formatting problem, to a video game, to mashups, and to linear programming.
18 November 2010
Speaker: Shuangyan Liu (Jenny), Department of Computer Science, University of Warwick
Title: Group Collaboration Problems and Causes
Abstract:
A majority of empirical studies show that learning through collaboration has brought a significant advantage over individual learning. Although empirical studies demonstrate the benefits that collaboration can provide for both online and face-to-face learning (e.g. better learning outcomes and enhanced group interactions), some recent studies have revealed that there are still many problems existing in group collaboration, which eventually affects the effectiveness of collaborative groups. Most of the problems have been induced by learners themselves. However, current literature fails to adequately address the factors, from the student perspective, causing such problems. In this talk, I present the group collaboration problems identified from existing literature and the results from a nationwide survey of university students toward their perceptions on the factors causing the problems. Last but not the least, I report the statistical results obtained from the survey for the correlation between student background and their perceptions of the causes.
11 November 2010
Speaker: Mike Joy and Russell Boyatt, Department of Computer Science, University of Warwick
Title: Introduction to the MALog Project
Abstract:
The MALog project is a three year project funded as part of the EuropeanUnion Lifelong Learning Programme. The primary aim of the MALog project is to provide high quality learning materials to help students and employees develop their expertise in mathematical logic. The project engages with high-schools, universities and industry to identify the learning material required and real-life examples of issues using mathematical logic. The project is a partnership between five institutions: the University of Warwick, Tampere University of Applied Sciences and Hervanta Upper Secondary School in Finland, and the Technical University of Civil Engineering and George Cosbuc National College in Bucharest.
This talk will serve is an introduction to MALog. We will discuss the results of a needs-analysis survey, the issues developing plans for units of learning materials and some of the technologies involved in producing learning materials. We will also discuss some of the challenges involved in the development of a mathematical logic ontology — a shared vocabulary and catalogue of mathematical terms. Finally, we will discuss the next two years of the project and the future research opportunities.
4 November 2010
Speaker: Meurig Beynon, Emeritus Reader, Department of Computer Science, University of Warwick
Title: Constructionist learning by computing for construal
Abstract:
A construal is a physical object that supports sense-making through exploratory interaction and interpretation. This notion was elaborated by the historian and philosopher of science David Gooding in his account of Faraday's experimental methods. The advent of computing technology has liberated the making of construals, but construal-by-computer is not well-served by the focus in computer science on principles and tools for developing programs.
In this talk, I shall present a prototype environment for creating construals by computer that exploits model-building with dependency such as is represented in spreadsheets and dynamic geometry environments. Our construals are made up of definitions that express dependencies between observables. As will be illustrated with reference to construing human solving of Sudoku puzzles, many kinds of human agency can be expressed through modifying the current set of definitions. The construal serves as a shared artefact with which developers, teachers and pupils can all interact concurrently in essentially the same way, each according to their role and experience. Preliminary experiments with schoolchildren highlight potential for rich and radically new kinds of learning experience and unprecedented scope for recording, monitoring and intervening in support of constructionist learning.
Background sources
- David Gooding, Experiment and the Making of Meaning, 1990
- The Empirical Modelling website:
http://www.dcs.warwick.ac.uk/modelling/ - EM publications on educational themes:
http://www2.warwick.ac.uk/fac/sci/dcs/research/em/applications/educationaltech/ - Workshops relating to a Sudoku solving construal:
http://www.dcs.warwick.ac.uk/~wmb/sudokuExperience/workshops/
28 October 2010
Speaker: Craig Stewart
Title: Personalised eLearning: Cultural Stereotypes
Abstract:
This talk addresses the research, development and use in the area of a learner’s cultural background as a basis for personalising their educational experience. Many modern eLearning systems use some degree of learner personalisation to adapt the delivery of content to better suit a learner’s individual preferences. The criteria for this adaptation vary from background knowledge, to learning styles. The cultural influence on a students ability and effectiveness to learn has had scant attention in recent years, which considering the spread of distance learning and the expanding multicultural nature of education is a severe flaw. The research to be presented in this talk addresses this lack of culturally motivated personalisation in eLearning by presenting a series of cultural stereotypes to be used in eLearning systems, specifically in Adaptive Education Hypermedia (AEH) systems. The results will aid both system developers and content creators in creating a culturally sensitive learning experience for their students.
21 October 2010
Speaker: Steve Matthews, Department of Computer Science, University of Warwick
Title: A Fuss about Nothing - toward a science of ignorance
Abstract:
'Information' in computer science is usually the irreversible accumulation of logically consistent totally defined so-called 'data'.
Yet in real-life information is often reversible (e.g. when a mistake is noticed), muddled up (e.g. concurrent), inconsistent (e.g. 'top' in multi-valued logic), and partially defined (e.g. denotational semantics). The more progress we make toward Mark Weiser's goal of the 'ubiquitous/pervasive' third generation computing the more we need to derive data from the actual structure of real-life. And so 'nothing' is chosen to be our working example for inter-disciplinary research in philosophy, logic, computing, & psychology having the ambitious goal of paving the way for a science of 'ignorance' to complement existing sciences of knowledge. This talk will discuss the speaker's approach to this research, presently being documented in an introductory book about 'nothing'. Topics covered will be (1) Why nothing matters (motivation), (2) Science and nothing (zero in mathematics, uncertainty principle in physics), (3) The logic of nothing (bottom and top in multi valued truth logic), (4) Computing nothing (incompleteness, partial information, negation as failure) (5) The spirituality of nothing (A case for God, death of God, Christian Atheism), (6) Nothing as ignorance (existentialism, non-self), (7) Nothing to live for (unconscious mind, depression, mindfulness) The conclusion will be that science and religion really do have 'nothing' in common.
14 October 2010
Speaker: Jason Nurse, graduating PhD student (viva talk), Computer Science, University of Warwick
Title: BOF4WSS: A to Z
Abstract:
When considering Web services' use for online business-to-business collaboration, security is a complicated and very topical issue. This is especially true with regard to reaching a level of security beyond the technological layer, that is supported and trusted by all businesses involved. Attempting to address these problems, my doctoral research proposes BOF4WSS, a Business-Oriented Framework for enhancing Web Services Security in e-business. The novelty and importance of BOF4WSS is its emphasis on a tool-supported development methodology, in which collaborating e-businesses could achieve an enhanced and more comprehensive security and trust solution for their services interactions. In this talk, I give an overview of the full framework (hence A to Z!) and go into detail on the specific problems investigated in my thesis. Finally, I report on the evaluations conducted and final thesis conclusions made.
1 July 2010
Speaker: Hugh Davis, Southampton University
Title: The Affordances of Web 2.0/3.0 for Personal Learning Environments
Abstract:
Traditionally learning has been seen as a solitary and individualistic task; learning has been represented as committing knowledge to memory and the personal acquisition of skills and literacies. The affordances of early computer technologies amplified this perspective, and transitions of learning technologies to networked platforms sustained the individualist context within the Virtual Learning Environment (VLE). However constructivist critiques of learning environments have emphasised the importance of social interactions and the benefits of groups working and problem solving as a means to learning and knowledge acquisition. Advances in Web technologies over the last decade (the so called Web 2.0) have enabled us to build tools to support and integrate many kinds of collaboration and learning in networks. Such tools have been retrofitted to existing VLEs.
This presentation argues that the current generation of Virtual Leaning Environments are no longer fit for purpose; they embody an approach to learning that supports ineffective/inappropriate didactic approaches, and do not complement the expectations or approaches to learning taken by Generation Y learners.
At the University of Southampton, as part of a major curriculum innovation project we have been examining the approaches we wish to take to providing the virtual environments for learning, teaching and research that will be fit for the next ten years. The focus of this presentation will be on our reflections on personal (personalised and personalisable) rich learning environments, and the part played in these environments by Web 2.0, linked data and cloud computing.
Biography:
Hugh Davis is Professor of Learning Technology within the School of Electronics and Computer Science, where he leads the Learning Societies Lab (http://www.lsl.ecs.soton.ac.uk
), a research team of 50-60 people with a focus on Web Technologies and Technology Enhanced Learning. Hugh has a long history of research in the Hypertext domain going back to before the advent of the Web, and has over 200 publications (http://www.ecs.soton.ac.uk/people/hcd/publications) in Hypertext and Learning Technology, as well as numerous grants in these areas. In addition to his research, Hugh is the University Director of Education responsible for technology enhanced learning (TEL) strategy, and this combination of activities enables him to take a research informed and leading edge perspective on the technology that can be delivered at a cross university level.
24 June 2010
Speaker: Ann Nicholson, Monash University, Australia
Title: Some Applications of Bayesian networks
Abstract:
Bayesian networks (BNs) have become a popular AI representation for reasoning under uncertainty, with successful applications in a range of domains. In this seminar, I will give a brief introduction to the syntax and semantics of Bayesian networks and illustrate their workings in a commericially available software package. I will present a general iterative incremental process for knowledge engineering Bayesian networks and then describe some BN applications I have worked on: an intelligent tutoring system for decimals, a BN for environmental risk assessment of native in the Goulburn Catchment, and a medical decision support tool, TakeHeart II.
Biography:
Ann Nicholson is visiting the Dept of Statistics at Warwick until the end of June. She is on sabbatical from the Clayton School of IT at Monash University, where she is an Associate Professor. She received her B.Sc (Hons) and M.Sc. degrees in Computer Science from the University of Melbourne. In 1992 she received her Ph.D. in Engineering from the University of Oxford, where she applied dynamic Bayesian networks to robot monitoring. After 2 years as a post-doctoral research fellow in Computer Science at Brown University, she took up a position at Monash University in 1994. Her areas of research interest are reasoning under uncertainty, Bayesian networks, knowledge engineering, decision support, user modelling, stochastic planning and monitoring, artificial life and evolutionary ethics.
24 May 2010
Speaker: Vincent Wade, Trinity College, Dublin
Title: Personalisation in the World Wide Web
Abstract:
Personalisation web research have been successful in researching personalisation in niche application areas where web content has been specifically designed for e.g. in Tourist Information Systems, eLearning and Museum systems. More recent years has seen “personalisation-lite” techniques being used by web search providers e.g. personalisation based on country of origin, or on comparison with similar or aggregate queries. However, personal use of the web extends far beyond just personalised content, and encompasses many dimensions of a web experience e.g. personalisation of tasks & activities, personalisation based on cultural preferences and values, and personalisation for social interaction and community engagement etc.
Next generation personalisation is tackling the issues of dynamic adaptation and composition of web content and services drawn from the heterogeneous open web. Such personalisation empowers the user via adaptive experiences combining traditional multimedia web content, end user generated content (wikis, ,blogs, forums, tweets, RSS feeds etc) as well as dynamically customising delivery to devices and multimodal interfaces etc.
In this seminar we will consider new directions and dimensions in personalised, adaptive web and how they can be addressed. We will investigate key challenges involving integrated open corpus & service personalisation, cultural adaptivity (including multilingual personalisation), and indicate how personalisation can enhance the web communities and the wisdom of the crowd. We will explore techniques and technology to enable multi-dimensional web personalisation i.e. personalisation of the web based on multiple, concurrent influences (from individual identity & personal properties, to context, device and service affordances).
Biography: Vincent is an Associate Professor in the School of Computer Science and Statistics, Trinity College Dublin . He is Director of the Intelligent Systems Discipline which is a cluster of research groups including the Knowledge and Data Engineering Group, the Graphic Vision and Visualisation Group, the Computational Linguistics Group and the Artificial Intelligence Group as well as the Centre for Health Informatics. Vincent is also Deputy Director of the newly formed Centre for Next Generational Localisation and Personalisation (CNGL) sponsored by Science Foundation Ireland. This multi university research centre is focused on the research and development of innovative digital content management, personalization and localization on the web.
See http://kdeg.cs.tcd.ie/ and http://www.cngl.ie/
20 May 2010
Speaker: Nugget Zemirline, SUPELEC, Paris, France
Title: Assisting in the Design of Adaptive Hypermedia Systems
Slides: Available as PDF
Abstract:
Adaptive web has emerged as a new challenge for the semantic web. One of the aims of the adaptive web is to adapt a set of web resources to users. Here, we address the problem of adaptation specification basing it on user characteristics. We present a pattern-based approach to express adaptation strategies in a semi-automatic and simple way. This allows the creator of an adaptive system to define elementary adaptations by using and instantiating adaptation patterns. These elementary adaptations can then be combined, allowing to specify adaptation strategies in an easy and flexible manner. We distinguish adaptive navigation according to two main criteria: the selection operations performed in order to obtain resources being proposed to the user and the elements of the domain model involved in the selection process. We present a taxonomy of elementary adaptive navigation techniques. An experiment with a group of volunteers shows that our approach is even more suitable than those based on ?traditional? AH languages.
20 May 2010
Speaker: Teemu H. Laine, School of Computing, University of Eastern Finland
Title: Technology integration in mobile-based learning enviroments
Slides: Available as PDF
Abstract:
Mobile-based learning is particularly suitable for informal learning activities, and it can be divided to three sub-types: mobile learning, pervasive learning and ubiquitous learning. Pervasive and ubiquitous learning spaces (PLS and ULS) create bridges from the real world to the virtual world, allowing the context-sensitive utilisation of real world objects and information in the learning process. We conclude that a model of technology integration for pervasive or ubiquitous learning does not exist. Based on a three year long process of mobile-based game development activities, we present foundations of such a model for game-based pervasive learning spaces. The model considers a diverse set of requirements from context, pedagogy and design perspectives, and it can be utilised by PLS designers and developers to determine which requirements must be considered when integrating technology into a PLS. To evaluate the suitability of the model for PLSs, we analyse two pervasive learning spaces with it. Additionally, we propose an evaluation scheme which can be used to measure the effectiveness of technology integration in a given PLS. Finally, future aspects of the model and our game development activities are discussed.
6 May 2010
Speaker: Andreas Loizides, Department of Computing, PA College, Larnaca, Cyprus
Title: The Empathic Visualisation Algorithm (EVA) and its Application on Real Data: The Financial and Fear of Public Speaking (FOPS) Cases
Slides: Available as PDF
Abstract:
This research work demonstrates a technique for visualizing multidimensional data sets holistically, and discusses results from applications in financial data as well as data gathered from a Virtual Reality (VR) experiment — Fear of Public Speaking (FOPS). The EVA algorithm to be presented provides an automatic mapping from semantically important features of the multi-variate data set to emotionally or perceptually significant features of a corresponding visual structure (such as a human face and more specifically human facial expressions). In other words a single glance at the visual structure informs the observer of the global state of the data, since the visual structure has an emotional impact on the observer. The visualization is designed to correspond to the impact that would have been generated had the observer been able to access the relevant semantics of the underlying data.
5 February 2010
Speaker: Dr David Bacigalupo, iSolutions/Learning Societies Lab, School of Electronics&Computer Science, University of Southampton
Title: A Formative eAssessment Co-Design Case Study
Paper: Available as PDF
Abstract:
Formative eAssessment can be very helpful in providing high quality higher education assignments. However, there are obstacles restricting the uptake of formative eAssessment in higher education including both cultural and technical issues. When a university is encouraging the uptake of formative eAssessment internally it is useful to have case studies from academic schools detailing how academics enthusiastic about formative eAssessment have used it in their modules. It is particularly helpful if these case studies document: /i.)/ the principle obstacles that these champions had to deal with; /ii.)/ a cooperative-design process through which these obstacles have been dealt with by the champions (with assistance from e.g. learning technologists); and /iii.)/ an evaluation of the effectiveness of the resulting formative eAssessments. However there is a shortage of such real-world long-term case studies. This paper helps fill this gap in the literature by describing the case of a Modern Languages module within a Russell Group university (Southampton). The formative eAssessment solution resulting from the case study utilises our QTI, mobile QTI, accessibility, and web 2.0 tools and can be positioned at the cutting edge of formative eAssessment practice. We have evaluated this with undergraduate student volunteers from Spanish modules and received positive feedback.
17 September 2009
Speaker: Toshio Okamoto, University of Electro-Communications, Toyko, Japan
Title: The organizational knowledge management for e-Learning practices in UEC and individual mentoring by Agents
Slides: Available as PDF
Abstract:
In this presentation, the concept of organizational knowledge circulated management system on e-Learning was given with the foundations such as organization al collaboration in a university, and facilities and contents in the center for developing e-Learning. As parts of the system, we have developed mentoring system and methods to construct effective contents from technological and pedagogical aspects, and evaluated the system with our e-Learning management
model. Then the important functions and evaluation aspects were described in order to go toward new information technology systems of e-Learning.
17 September 2009
Speaker: Mizue Kayama, Shinshu University, Japan
Title: Skill Learning Support : A Basic Study on a Drawing-Learning Support System in the Networked Environment
Slides: Available as PDF
Abstract:
The purpose of this study is to develop a support system in drawing-learning within a networked environment. In this presentation, we describe the results of potential assessment for our system. Two assessment approaches are shown. One is the possibility of a digital pen as a drawing-tool. The other approach is the effectiveness of the drawing-learning support in the networked environment, based on the reuse of the learner's and/or expert's drawing process. The drawing process model for supporting individual drawing-learning is also discussed.
1 May 2009
Speaker: Robin Burke, DePaul University, Chicago
Title: Robust Recommender Systems
Abstract:
The openness and anonymity of the Internet environment create many hazards for e-commerce. For collaborative recommender systems, it raises the possibility of that attackers will seek to bias the output recommendations through manipulation of the public inputs that the system permits. Fighting such manipulation is a constant battle for the owners and maintainers of such systems. In this talk, I will describe the known vulnerabilities of collaborative algorithms and examine a range of possible attack types that could be deployed against them. With these vulnerabilities in mind, I will discuss possible responses, including the deployment of alternate recommendation algorithms and the use of supervised and unsupervised techniques to detect attacks. Building on this research, I will examine what it might mean to build a robust collaborative recommender and consider the implications for other machine learning techniques deployed in public on-line environments.
30 April 2009
Speaker: Sarah Lim Choi Keung, Warwick University
Title: Building a Trust-based Social Agent Network
Abstract:
Agents evolving in a multi-agent system interact with one another to achieve their individual goals. In trust-based models, agents form a local view of their environment from their direct interactions, and base their interaction decisions on the trustworthiness of the other agents. Agents can also obtain recommendations about other agents from third parties, either directly or indirectly. While trust and reputation ensure that an agent selects and interacts with the most appropriate provider, the agent can also learn about agent interconnections in its environment. By building a network of agents, and with information about interaction details, trustworthiness, recommendation chain and reputation, the agent is in a better position to extract emergent information, such as potential new customers, suppliers, its competitors and potentially collusive groups of agents. In this talk I will present a mechanism for agents to build a representation of their local environment based on direct interactions, trust and reputation.
This is a practice run of a workshop presentation, so attendance and feedback will be appreciated.
20 March 2009
Speaker: Bettina Berendt, Katholieke Universiteit Leuven, Belgium
Title: Sailing the Corpus Sea: Tools for Visual Discovery of Stories in Blogs and News
Slides: Available as PDF
Abstract:
Rich information spaces like blogs or news are full of "stories": sets of statements that evolve over time, made in fastgrowing streams of documents. Even if one reads a specific source every day and/or subscribes to a selection of feeds, one may easily lose track; in addition, it is difficult to reconstruct a story already in the past. In this talk, I present the STORIES methods and tool for (a) learning an abstracted story representation from a collection of time-indexed documents; (b) visualizing it in a way that encourages users to interact and explore in order to discover temporal "story stages" depending on their interests; (c) supporting the search for documents and facts that pertain to the user-constructed story stages; and (d) navigating in document space along multiple meaningful dimensions of document similarity and relatedness. This combination provides users with more control, progressing from "surfing" the (Social) Web to "sailing" selected corpora of it, semantically in story space as well as between the underlying documents. An evaluation demonstrates that machine learning and interaction lead to representations that serve to retrieve coherent and relevant document subsets and that help users learn facts about the story.
19 March 2009
Speaker: Gbolahan Williams
Title: Harnessing the Web for Improved Recommendations
Abstract:
The proliferation of opinion within blogs and review web sites has over the last decade increased interest in the development of automated techniques for generating sentimental recommendations. The key elements for this process include document retrieval through crawling of web documents; document classification & categorization of the crawled documents into review and non-review documents; sentiment analysis & modelling based on the classified set of documents and event recommendations based on event sentiment profiles. I will discuss my masters research which looked at these individual processes. Drawing on empirical tests carried out in our work, we believe that there are still a number of unanswered research questions related to the reliable extraction of sentiment associated with reviews, some of which we highlight within our work.
3 December 2008
Speaker: Michele Gorgoglione and Sabrina Lombardi
Title: Context in Personalization: Experimental Analysis on Customer Modelling and Recommender Systems
13 October 2008
Speaker: Teemu Laine
Title: Pervasive Learning Environments
Slides: Available as PDF
15 July, 2008
Speaker: Nuno Escudeiro, Universidade do Porto
Title: Focused semi-automatic content retrieval: learning from incomplete specifications
Abstract:
Organizing large collections of items on a set of classes requires previous work on labeling cases to train a classifier. Case labeling is expensive and, besides that, retrieving cases from under represented classes may be very demanding. These problems arise frequently when building web resources, working with text collections or any unstructured corpus. Active learning attempts to reduce the amount of labeling by criteriously selecting cases and querying an oracle to label them. We propose a new active learning algorithm, d-confidence, aggregating distance and confidence. We show empirically that our approach outperforms traditional active learning achieving faster coverage of case space while improving error rate
15 July, 2008
Speaker: Tao Li
Title: Automated Taxonomy Generation for Summarizing Multi-type Relational Datasets
Abstract:
Taxonomy construction provides an efficient navigating and browsing mechanism to people by organizing large amounts of information into a small number of hierarchical clusters. Compared with manually editing taxonomies, Automated Taxonomy Generation has numerous advantages and has therefore been applied to categorize document collections. However, the utility of this technique to organize and represent relational datasets has not been investigated, because of its unaffordable computational complexity. We propose a new ATG method based on the relational clustering framework DIVA. By incorporating the idea of Representative Objects, the computational complexity can be greatly reduced. Moreover, we analyze the divergence of the data attributes and label the taxonomic nodes accordingly. The quality of the derived taxonomy is quantitatively evaluated by a synthesized criterion that considers both the intra-node homogeneity and inter-node heterogeneity. Theoretical analysis and experimental results prove that our approach is comparably effective and more efficient than other ATG algorithms.
22 April 2008
Speaker: Helen Ashman
Title: The Emergence of Semantics on the Web
Abstract:
Many researchers aim to add semantics to the Web by supplementing Web resources with metadata describing the resource. The metadata is generally added by the author/owner of the resource and reflects their personal judgement. We take an alternative approach which is to mine the semantics expressed by the consensus of user interaction with search engines, derived over time and from many users. These semantics are an emergent property of the mass of user interactions with all of the search engines, the "collective wisdom of crowds", and are neither generated nor controlled/owned by any entity.
Ongoing work includes the ad hoc identification of reliable semantics at an early stage, rather than the usual post hoc identification applied in Information Retrieval algorithms. We aim to identify emergent semantics before they validated by a statistically significant sample, by condensing the characteristics of emergent semantics from the early stages of the clustering which forms them. This is partly motivated by the sparsity of available data - much data exists but privacy concerns mean that it is not readily released even for research purposes. By identifying emergent semantics at an early stage, the productivity of a limited supply of data can be maximised.
By understanding the characteristics of emergent Web semantics it is not only possibly to identify clusters as early as possible, but also to predict which "proto-clusters" will coalesce into genuine semantic clusters. This work contributes to an overarching framework of emergence formalisation, being a specific application area wherein emergence can be detected and manipulated.
30 October 2007
Speaker: Tao Li
Title: DIVA: A Variance-based Clustering Approach for Multi-type Relational Data
Abstract:
Clustering is a common technique used to extract knowledge from a dataset in unsupervised learning. In contrast to classical propositional approaches that only focus on simple and flat datasets, relational clustering can handle multitype interrelated data objects directly and adopt semantic information hidden in the linkage structure to improve the clustering result. However, exploring linkage information will greatly reduce the scalability of relational clustering. Moreover, some characteristics of vector data space utilized to accelerate the propositional clustering procedure are no longer valid in relational data space. These two disadvantages restrain the relational clustering techniques from being applied to very large datasets or in time-critical tasks, such as online recommender systems. In this paper we propose a new variance-based clustering algorithm to address the above difficulties. Our algorithm combines the advantages of divisive and agglomerative clustering paradigms to improve he quality of cluster results. By adopting the idea of Representative Object, it can be executed with linear time complexity. Experimental results show our algorithm achieves high accuracy, efficiency and robustness in comparison with some well-known relational clustering approaches.
21 June 2007
Speaker: Gabriel-Miro Muntean
Title: Wired and Wireless Quality-Oriented Adaptive Multimedia Streaming
Slides: Available as PDF
21 June 2007
Speaker: Cristina Hava Muntean
Title: Performance-aware Adaptive E-Learning Environments
Slides: Available as PDF
21 June 2007
Speaker: Maurice Hendrix
Title: Evaluating Adaptive authoring of Adaptive Hypermedia
3 April 2007
Speaker:Lorna Uden
Title:Collaborative Design Using Activity Theory
Slides: Available as PowerPoint
Abstract:
Email is the most common collaborative tool in use today. Although originally designed as an asynchronous communication tool, it is being used increasingly for information management, coordination and collaboration tasks. For effective collaborative work, email must be designed that meets users' needs and their experience. The traditional approach to designing interfaces has been increasingly criticised because of the gaps between research results and practical design, especially concerning requirements. Requirements elicitation is a key to the success of the development of all email applications.
Activity theory incorporates the notions of intentionality, history, mediation, motivation, understanding, culture and community into design. In particular, it provides a framework in which the critical issue of context can be taken into account. This paper describes the use of activity theory for the requirements analysis of a collaborative email system for a manufacturing company, XBC Ltd.
9 February 2007
Speaker: Maurice Hendrix
Slides: Available as PowerPoint
Title: Evaluation of Adaptive Authoring of Adaptive Hypermedia
Abstract:
Now-a-days much research exists into the design and implementation of adaptive systems, but still, much less exists that tackles the complex task of authoring of such systems, or their evaluation. Manual annotation has proven to be the bottleneck of authoring of adaptive hypermedia. Therefore, all means for supporting this authoring process by reusing automatically generated metadata are helpful. Previously we have proposed the integration of the generic Adaptive Hypermedia authoring environment MOT into a semantic desktop environment. A prototype based was built upon this integration approach. This paper describes an evaluation of both the approach in general and the prototype in particular.
28 November 2006
Speaker: Tao Li and Sarabjot Singh Anand
Title: Knowing Me, Knowing You: The State of Play with Recommendation
Abstract:
Recommender Systems play an important role in mediating between users and the information overload presented by the Internet, by attempting to know what the user wants without the user having to explicitly state what he is looking for. In this talk we first describe the current state-of-the-art in Recommender Systems, starting with what data is collected on users browsing the Internet, the web mining techniques that have been used, the user models that are build, how they are used to ultimately generate recommendations and how these systems are evaluated. We then describe the specific areas of this space that our current research is contributing to, what has been achieved, what still needs done and how we plan to do it. We will end the talk with a brief description of the current challenges facing the recommender systems community at large.
7 November 2006
Speaker: Alexandra Cristea
Title: Adaptive Hypermedia: What is it and Why are we doing it?
Slides: Available as PowerPoint
Abstract:
There is a vast diversity of learners in the world; each person has their own learning preference, aims and objectives. Every learner has the right to demand a high quality, personal learning experience. However as current web-based learning environments offer a “one-size-fits-all” approach to the delivery of learning materials (every learner is given the same set of resources), the personalised approach to education is sadly lacking from most online systems. Adaptive Educational Hypermedia (AEH) seeks to address this lack. It aims to create new opportunities for learners, whilst also enhancing existing approache - delivering lessons and courses adapted to the requirements of each learner.
Therefore, Adaptive Hypermedia (AH) (Brusilovsky, 2001) can be considered the solution to the problems arising from the “one-size-fits-all” approach to information delivery prevalent throughout the WWW today.
The way Adaptive Educational Hypermedia (Brusilovsky, 2001) caters to the needs of each individual student is to adapt to their goals (Clifford, 2000); knowledge level (De Bra, & Calvi, 1998); background; interests (Brusilovsky et al., 1996); preferences (Hak et al., 1997); stereotypes (Zakaria & Brailsford, 2002); cognitive preferences (Chen & Macredie, 2002) and learning styles (Stach et al., 2004).
Today there are a few such AEH systems, many of which are as yet experimental (e.g., Interbook, AHA!, TANGOW, WHURLE). Their focus is on the learner”™s requirements: they adapt their educational content to different dimensions of each learner, such as: knowledge levels, user goals, tasks, educational context (e.g., are they in school, university, or learning from home?), and more recently learning styles (e.g., LSAS).
24 October 2006
Speaker: Steve Russ and Meurig Beynon and Antony Harfield
Title: Modelling and Learning
Abstract:
The high expectations of the 1980's and 90's for the impact of computers and technology on education and learning have been realised, if at all, only very slowly. Some responses to this situation emphasise the cultural and contextual problems in enabling technology to mediate rich and flexible learning experiences to students (for example, negative attitudes of teachers or technical limitations of equipment). A complementary, but less common, approach is to develop alternative ways of using computers and alternative ways of developing software that allow the learner to build or modify artefacts in open interactive environments rather than responding in pre-conceived ways to a conventional `program'. Such environments allow the learner to develop personal models of a domain in an exploratory and explanatory fashion. This is one application area of the wide-ranging approach to computing being developed in the Empirical Modelling (EM) group. The emphasis in EM is to use the computer to embody and visualise our understanding of a domain before conceiving of any function or requirement for a program. Theoretically, this has led us to give priority to the `meaning-making' concepts of observables, dependency and agency (rather than data structures and algorithms). Practically, it has led to tools with which numerous students have produced artefacts that are better described as conceptual or cognitive models, rather than programs, and are well-suited to learning. The seminar will give a brief overview of the EM perspective and, if possible, its relationship to surrounding fields. Several recent and relevant papers are available from the `Publications' tab at empiricalmodelling.com.
28 September 2006
Speaker: Sarah Lim Choi Keung and Nathan Griffiths
Title: Agent-Based Systems: Trust, Reputation, and Social Relationships
Slides: Parts 1&3 (pdf), Part 2 (ppt)
Abstract:
Agents are entities that perceive their environment and act autonomously and pro-actively, in order to solve particular problems, whilst remaining responsive to the environment. Using agents allows us to view computation as interaction, with applications being built from societies of distributed components. Agent technology has been widely applied to diverse domains including e-business, telecoms, web, and service-oriented architectures.
Uncertainty arises whenever agents interact; the autonomy of individuals can lead to failures of cooperation, or unexpected outcomes. Uncertainties may result from unreliability, competition, malicious behaviour, and from the open nature of many multiagent systems. Agents can use knowledge of inter-agent relationships, trust and reputation, and coalition membership to cope with uncertainty and enhance their performance.
This talk introduces agent technology in general, and gives an overview of the research areas of trust, reputation, social relationships, and coalition formation. Finally, we give an overview of our current research activities.