Beccy and I recently attended the European Conference on Research and Advanced Technology for Digital Libraries (ECDL) in Glasgow. We were there partly to present a poster about the LIFE-SHARE Project and partly to get up to date with developments in the field. The conference covered a range of topics, some of which were a little too ‘advanced’ for us non-computer science specialists! However, many streams were of relevance and interest. Here are a few of the things which I took away.
- Susan Dumais from Microsoft Research, in her keynote speech, talked about how existing information retrieval tools (search engines and the like) treat the internet as a static environment, in which content never changes. In reality, however, web pages are constantly changing. Different types of pages have different change patterns over time. Different types of page also have different revisit patterns, depending on the type of content they have and how people use that content (compare a news homepage which might be visited very frequently in a short space of time as a user engages in ‘hub and spoke’ navigation with an email homepage, which a user may visit regularly once a day). Dumais presented two useful tools: DiffIE, a browser plug-in which highlights web page content which has changed since last time you visited that site; and TemporalIR, a search ranking tool which looks at how often a web page changes (there is a correlation between how frequently a website is updated and it’s quality) and the longevity of terms within it (if a webpage always has a certain term in it, it is likely to be more relevant and less transient). This can help improve the relevance and quality of search results. One consequence: web content providers will have to make sure key pages are updated frequently to get high up in search engine results. A copy of Susan Dumais’ PowerPoint slides for the talk can be found here.
- Work on the privacy implications of folksonomies by a team from Karlsruhe Institute of Technology was thought-provoking. They outlined how personal information (about interests, opinions and activities etc.) can be automatically gleaned from the tags users create on sites such as Flickr and Last.fm. Normally, the value of such social media is that they are about sharing, but the authors make the point (and this prompted an animated discussion in the question-time) that users may not always want everything they tag to be public. They may intend it for their own social circle only, or for their own use (I can concur with this feeling, having abandoned Delicious for the more private Bookmax). The team have developed a method of encrypting personal data and enabling it to be kept private, even from the website owners. However, it relies on web site owners implementing it, when they might have vested interests in the information contained in their users’ tags (for example to target advertising). The incentive to use the method would have to come from users putting pressure on web site providers.
- Seaweed (seamless web editing) is a framework for editing websites directly without the need to switch to a separate editing mode, common in wikis and blogs. This looks potentially very useful, as a way of making the editing process more direct and less fiddly.
- Sandra Toze and Elaine G. Toms’ study of group work was very interesting. They observed groups of students working on real assignments, using video/audio feeds, computer log files and diaries kept by the students themselves as the sources of data. The interesting thing was to observe how complex group work actually is, with the students going through different phases of working, splitting up to work separately and then joining back together again. The range of tasks they carry out includes administrative tasks, communication, information seeking and retrieval and creation. The students did not always work in the most efficient way, for example failing to find important resources until late in the project and then having to revisit earlier stages in order to incorporate new information. The students also used the space they were given to work in in ways that supported this complex pattern of work, with the group often breaking into sub-groups to gather round different PCs, or individuals breaking off from the group to work on their own. The group would then often reconvene and arrange the space so that they could all work together again. Toze and Toms discussed how digital libraries can support group work, with reference to Marchionini’s concept of the ‘sharium’. The digital library was just one of a number of tools which the students used for different purposes and at different stages. There is a need to support not just information retrieval but also working with information – bringing diverse pieces of information together and build them into interim and then final information objects. The digital library needs to integrate collaborative tools.
- CritSpace is an interesting tool being developed by TEES Center for the Study of Digital Libraries at Texas A&M University. It is based on the premise that researchers have a particular need in the early, formative stages of their research to express “ambiguous, partial and emerging knowledge-structures”. The tool allows resources from digital repositories to be brought together into a workspace and arranged visually. There they can be annotated and a viewer allows them to be zoomed into. It is interesting that this platform takes into account how the creative process works in academic (and any other) work.
The poster presentation went well and we received quite a lot of interest. The conference was hosted by the University of Glasgow, in the Boyd-Orr building located in the West End, a stone’s throw away from Byres Road. As Beccy and I both studied in Glasgow, it was a nice opportunity to revisit the Glasgow experience (including the distinctive square sausages, which we bought every morning at a van!)