Open Impact

November has been one of the busiest months we have ever experienced in our interactions with business and industry!

In the week that Lewis Hamilton won the F1 Drivers and Constructors’ Championship for Mercedes AMG, the company’s Head of Electronics is giving a talk to our student branch of the IEEE. Only a few days ago, third-year Electronics student Josh Oldfield was attending the National Microelectronics Institute annual dinner as runner-up in the UK Electronic Skills Foundation ‘Scholar of the Year’ competition.

ARM and Imagination Technologies have both held fast-track interview sessions in the School and companies visiting ECS to give presentations or provide support to students this month include NCC, J P Morgan, New Voice Media, FactSet, Goldman Sachs, Accenture, Bloomberg, Entrepreneur First, MLH, QinetiQ, Halma, Cundall, Thales, IBM, National Instruments, and ECTalent. New companies joining the Careers Hub this month include Acano, Babcock, Davide Leone & Partners, Haseltine Lake, Oxford Computer Consultants, Sparx, TE Connectivity, and Unboxed Consulting. We now have over 140 companies on our Affiliated webpage, and over 60 companies are booked to attend our annual Engineering and Technology Careers Fair on 10 February.

Our employability figures for ECS graduates in 2013 (the latest available) reflect the high degree of interest that employers have in ECS students. ECS Electronics and Electrical Engineering graduates had an employment rate of 96 per cent six months after graduating in 2013, and Computer Science and IT graduates 95 per cent. These are outstanding figures and place us amongst the top five universities in the UK for our subjects.

We continue to work very hard with our partner companies to create events and opportunities that are of value to them as well as to our students, and to help ensure that our students are well informed about career choices, and well prepared for the application and selection process. Students (especially since the advent of tuition fees) are more motivated to explore career opportunities from the earliest days of their degree courses and keen to undertake summer internships after their first year. Part of our offer to prospective students centres on our excellent links with leading companies, and the strong relationships we already have with our Laureate and Affiliated companies, as well as with sector partnerships such as the Power Academy and the UK Electronic Skills Foundation.

November is, of course, a peak time for careers, since many organizations have their closing date for graduate applications over the next few weeks. But  careers are not just for Christmas! and with our Careers Fair fast approaching on 10 February and then the next round of applications for summer internships opening up, careers and employability is now very much a year-round activity!

For further information about the ECS Careers Hub, contact Joyce Lewis;  tel.+44(0)23 8059 5453.

I demonstrated the Timeline work to the University’s Research and Innovation office today; they were so excited by its possibility that they dragged the DVC for Research out of his office and showed him. Now he wants me to demonstrate it to all the School Heads of Research. It just emphasises the impact that a well chosen (tiny) piece of technological innovation can have!

There is now an EPrints export plugin that produces a suitable RSS feed for creating a timeline. It had to be changed to (a) list more than the default (10) number of recent items, potentially listing the whole repository if necessary and (b) report items by their date of publication in the literature, not date of deposit in the repository. This last point is one of those interesting departures from the normal Web model that a repository lives with: its resources are not Web pages, they are papers or articles that have a role in “the scholarly literature” as well as the “academic web”. This feed reports the “latest things in the literature that feature in the repository” not “the latest things in the repository”.

This means that we can easily create a timeline based on data from the repository (e.g. all papers on topic X which have more than 10 citations); once in the TimeLine editor, it is easy to tweek the individual items.

Since timelines seem to be a key artefact for helping external parties (e.g. journalists) understand the gradual buildup of research contributions in an area, it seems appropriate that we should devote some resource to helping make timelines.

To this end we are looking at creating appropriate feeds from our three main school databases (eprints, projects and press releases) so that each will produce an ATOM feed of formatted entries corresponding to searches on a particular topic. These can then be used to automatically create a timeline, which can then be edited in the TimeLine desktop application for clarity.

During our recent progress meeting we were discussing the kind of material that we would need to pull together to create a “research case study”. We took “Agent Research” as the exemplar, because there is a decade of work in our school that comes under that heading. When we described the wealth of material available – papers in the repository, press releases going back to 2004 and and entries in the projects database – the journalist with whom we are working remarked that she would need a timeline to be able to see the contributions of all these projects as they had built up over the last ten years.

We looked at each other and thought – of course! A timeline – how obvious! The problem is that timelines are not very easy to create (Excel), or else the software that is available is not very intuitive (Similie) or attractive (Similie and Excel). Eventually I found an application called “TimeLine 3D” for the Mac. It produces very pretty looking timelines, with useful 3D renderings, and (even better) it accepts input from RSS/Atom feeds and it can be scripted.

To get the feel for it, I have created timelines for three pieces of extended research: Agents, Web Science and Repositories. The first mainly uses papers from the repository (the three most highly cited from each year), the second mainly uses press releases and the third project descriptions (and software developements). The first is a summary of significant activity, the third is an attempt to comprehensively cover the field. We are not sure yet which makes a good timeline for this purpose.

See: Agents Timeline, WebScience Timeline, EPrints Timeline

Useful meeting with the ECS communications team to discuss the creation of some exemplar Impact case studies based around the research in ECS. Influenced by the rumblings slowly emerging concerning the nature of REF impact statements but also looking at the broader picture, we have identified one or two good stories to run with and they are being worked up as I type.

As well as generating the case studies we will also be document the process of generating the case studies to continue to identify areas where we can improve the process of initial data capture about research, help agregate research data to facilitate the initial construction of case studies and also try to identify additional data sources that it may prove beneficial to try and capture at source rather than after the fact.

All told, a productive meeting and I’m looking forward to seeing the first draft of a case study emerge in the next couple of weeks.

These are the notes from the Open Impact project Kick Off meeting which took place on 18/1/10

Introductions

Mark Weal – lecturer, ECS. Interested in telling stories (narratives) about research.

Joyce Lewis – communications manager, ECS.

Hugh Davis – head of learning societies lab, ECS.

Colin Smith – Repository manager at the Open University. Interest in the project but not sure how to get involved at the moment. Interested in how repositories can gather impact information. Store impact alongside the research it is about.

Bill Mitchell – Director of BCS Academy of Computing to cater for academics. Public engagement with academic aspects of computing. Making the wider audience aware of value of real computing research.

Sarah Chaytor UCL. UCL developing institutional research information service.

Wendy White – Manager of institutional repository and part of REF team. Can it be used for collecting information for the REF. Improving interdisciplinary impact.

Robin Axford – Research and Innovation Services, Southampton. Faciliating the research lifecycle project. Looks at whole lifecycle including outputs and impact.

Les Carr – Lecturer in ECS and EPrints repository platform and development. Open access push. Want to use repositories for everything. or, trying to span the gap between individuals and the Web. Encourage people to document information and make it available so it can be useful to a wider community. Take things that are happening to individuals that build up over time and make them available to other communities.

Nick Gibbins – lecturer in ECS. Organisational knowledge management using Semantic Web. Gathering information about projects and research in the UK on a previous RI project (dot ac).

Chris Gutteridge – lead developer on EPrints. Head of Web for ECS.

Dave Challis – Webmaster for ECS.

Patrick McSweeney – developer (ECS)

Marcus Ramsden – developer (ECS).

Discussions based on project proposal:

Bill – EPSRC and RCUK want the BCS Academy to push the impact agenda.

Joyce/Robin – Not clear University is engaging with the learned societies.

Sarah – links at an individual level but not at University level

Bill – Often work at a departmental level

Les – Are we going out on a limb by working on this relationship

Robin – Not sure that is the right question. Growing importance of multidisciplinary approach may not sit comfortably with learned societies.

Sarah – Not getting drawn down a path of definitions of impact and letting institutions tell the stories about what they want.

Bill – Not interested in gathering the numbers.

Wendy  – Still exploring the impact and ref. Complexity is coming out in the pilot.Some disciplines will want to tell a twenty year story. Moving between institutions becomes a big issue with this. This project could help contribute to that debate.

Colin – The criteria is very specific so the actual impact has to have happened in a very short window. 2-3 year window. Very specific thing to capture. Tories have said they may well suspend the REF for two years and abandon the  impact element.

Les – When we put this together we focussed on the story part of the impact rather than the numbers and measures.

Les – Are there parts of the impact indicators document we should be using.

Sarah – Yes, we should try and use them.

Colin – Focus on the collective rather than the individual. When we think in terms of repositories, linking up individual outcomes into clusters and gathering impact on that.

Chris – Could identify groups after the fact.

Wendy – collaborative working, not just a REF thing. Will be fewer bigger bids in the future.

Chris – secondary impact could be important also.

Wendy – Internal and external interfaces. Preservation can be important to help maintain that story.

Les – There becomes a point where metrics become a story.

– David Schulenberger’s closing keynote at SPARC DL meeting (Nov 2008): the job of the institutional repository is to tell the story of “what we’ve achieved” to its faculty and its institution’s funders and supporters. People who don’t want to put stuff in repositories often don’t want to engage with the public, or other audiences.

Going through the three use cases:

1) Stuff not being directly to hand or available. Collecting stuff together easily.

Developing case studies – considered, analytical, thorough.

2) Gathering information that may be useful for audit or proving claims of impact. Using lightweight tools to quickly capture information. Get clues about what is happening and capturing that.

3) Need to be able to create presentations easily from the high impact information.

New England journal of medicine, automatically creates slide sets about the research.

4) Looking at economic factors and interpretations. Can we make this simpler. Might get this information from bids. Can provide chain of evidence. Need to present an information context.

5) Developing case studies, taking a retrospective approach.

Joyce – all about timescales.

Bill – simple stories telling how things have changed, backed up by facts. Want to engage people with what is going on in the background by telling stories and give them clear evidence of how things have changed.

For industrial collaborators we want them to read our material and be enticed to give us lots of money. Want a track record to show how our technologies have had an impact. Anecdotal is good. And linked back into the chain of evidence. Emerging story, emerging impacts.

Not creating research journalists but to help the process along.

Forensic impact analysis. Be able to respond quicker to trends. Want to capture public engagement, conversation with industrialists, etc.  notability.

Linking repositories of other institutions that are collaborators on projects.

Might need a use case for someone putting together a case study of impact for REF.

Starts with outlining the research outcomes intended from the research.

Need to keep it very basic. Perhaps improve the bid storing mechanism.

Limit to the communications team know how and also academics struggling to communicate their research to the communications team.

Might be useful to provide a user friendly model for academics to desrcibe their research in impact terms.

Colin: At OU making use of request copy function to use as anecdotal evidence of impact. Bit of a competition between the open access policy or asking people to request information.

Bill talks about the BCS Impact repository for people to put research outputs in and be stored nationally. Good from a society point of view because the stories are from a lot of University.

Les: What are going to be in your repositories?

Bill: The case studies. Not meant to be prescriptive but we will try and engage people to get them to present their case studies as a story. Often get cold called for experts in particular area so we could use the repository to find experts.

Colin: Not sure how keen people would be in sharing impact case studies.

Discussing about how the tools might be trialled in OU and UCL. Both are keen to try and help.

Les/Robin: Potential for using a model of impact development to drive the process of impact collection of an extended period of time. E.g. Robin’s melon visualisation (to be discussed further).

Summary

The key issue seems to be collecting information that allows press releases and research case studies to be prepared.

Press release: timely, attention grabbing. Provides stories of breakthroughs or activities.

Research case study: considered, analytical, thorough. Provides evidence of impact.

Evidence may come from external sources (news, journals), as well as internal (CRIS, background section of research bids). It may be scraped from newspapers, websites and blogs in the form of ‘forensic impact analysis’. Notability and trending should not be underestimated as ways of alerting researchers to their instantaneous impact. Impact is an ongoing story, inevitably tied to the ebb and flow of current issues and interests in the public sphere. We should engage in ‘impact management’, similarly to ‘search engine optimization’.

Reworked work/cases

Create, in conjunction with BCS, a national ‘high impact CS research’ repository for depositing ‘research case studies’ (which may be simply powerpoints of presentations about a group’s expertise, or it may be structured and argued analysis of extended research impact in a field).

Make tools to help researchers deposit evidence of impact

Make tools to help comms managers pull together the evidence to create press releases and research case studies

Discussion Whiteboard Images

Subsequent BCS Repository Discussions

Southampton to produce a pilot ‘national subject high-impact research repository’ for BCS, which will be transferred to the BCS for hosting/management post-April.

BCS repository will collect ‘research case studies’ (RCS) of high impact research (HIR). Exactly what constitutes HIR will be decided by the institutions who submit the work; exactly what constitutes an RCS will evolve over time. Initially it may just be a publicity PowerPoint presentation (as suggested by Robin Axford) or it may be an analytical report (as suggested by Sarah Chaytor).

The repository is NOT collecting research outputs (it would have a hideous copyright position). Nor is it highlighting individuals or institutions. It exists to promote computing research.

The repository will be configured such that its deposits are of type ‘Research Case Study’ with the usual repository bibliographic metadata recast to be appropriate to this type. For example,  there must be a clear distinction between ‘author/creator’ of the RCS and the authors/creators of the HIR.

The repository must contain sufficient metadata to allow the BCS to produce various reports and visualisations for its stakeholders, including lists of (and links to) individuals, teams, organisations, projects and publications. It will not maintain authoritative lists of entities (people etc), but it will expose information suitable to engage in Linked Data applications.

The repository may contain information provided about individuals (specifically: photographs and brief bios) but it will not operate a ‘user profile’ service, maintaining up-to-date career histories, biographies and CVs of individuals. Such a service may be a future extension that the learned society may wish to offer, potentially as a subscription.

Proposed Metadata

Basic bibliographic metadata re deposited document (RCS) plus

• Keywords / tags
• What’s the story? (abstract of RCS, the interesting part extracted., the elevator pitch)
• Who was involved
• What are the implications for ME & what are the implications for SOCIETY
  • For a CEO this may mean ‘what are the implications for my business’ and ‘how will this change the market in which I operate?’
  • For a JOURNALIST this may mean ‘give me facts, analyses and contexts’
  • For the public this may mean ‘how will it change my life and my family’s life’ and ‘how will it change society’
• Where do I go next: if I’m interested, whom do I contact (this may be a research office, or an innovation and business consultant)
• Further reading: links to the University’s own pages and promotion on this work. May be press packs. Should contain more detail about the work, but not in the BCS domain or under their control.

The following items are key information for reports, visualisations, mashups and the Linked Data agenda.

• Key Individuals
• Name, affiliation, URL. Poss photo and brief bio.
• Key Projects
• Key Publications
• Key Partners

The RCS may describe work from many people, interdisciplinary across research teams, collaborative across research institutions and partnered with a number of commercial and industrial concerns.

Welcome to the Open Impact blog, which tracks the activity of the JISC “Open Impact” Rapid Innovation project.

Open Impact is a project to help collect evidence about the impact of research that has been undertaken in UK universities and to provide it to a range of stakeholders (government, funders, press etc) through an independent third party agency (a learned society). The project focuses on a specific discipline (Computer Science) mediated through a particular society (the British Computer Society).

In particular, this project will produce software that helps to make institutional repositories effective in collecting evidence of the impact of their institutions’ research – evidence that justifies the investment that government and research funders have made and that promotes the role of Universities in society.