Serendipity - Finding things you don't know you want!

March 03 2008

For the past couple of months, I've been looking into how you find items that you didn't know you wanted until you've seen them. Serendipitous finds or random browsing, if you will. This is a natural activity in a library, but methods need to be developed to expose the dark corners of a digital repository. Adopting some Web 2.0 sugar, I'm now working on publishing keywords from content in the repository as tag clouds so that at the bottom of the search page, people can get a feel for what are the most common (the most recent, the most popular, the most searched for) keywords in the repository.

The conference in Edinburgh helped a great deal. I've picked up snippets of code from Martin Morrey and Paul Hart that will help with trawling the repository using what's called OAI harvesting. The other comment that reverberated at the conference is making things easy to use. "Make it like Google or Amazon" is the challenge, nothing terribly new, just making the user experience better, more enjoyable. If I can make the repository a popular resource, I feel I'll have made some progress towards that goal.

Keywords: browsing, google, repository, searching, tag cloud

Posted by Boyd Duffee|

Learning Development Team

February 27th 2008

The Learning Development Team (LDT) in the Faculty of Humanities and Social Sciences is currently involved with supporting staff as they redesign modules for the new Keele Honours. Incorporating study skills into these modules is an important factor and the LDT wanted somewhere to be able to share existing resources with academic staff. The hope is that this will inspire staff, as well as encourage sharing and repurposing of objects.

Using the workflow provided, I have started adding resources with some basic metadata, and all including the tag “Faculty First” – the name of the project. This will allow staff to search for this tag and find all the material relating to first year skills in HumSS. Adding materials has been straight forward and allows instant access via the WebCT powerlink. The main problem I’ve faced is that the search function of the powerlink isn’t working as expected. Users cannot do a Google type search at the moment as only phrase searching is working. Intrallect are aware of this problem and are working to resolve the issue. As soon as this is sorted, I’ll be able to send out instructions to academic staff so they can view the materials and provide feedback to me on how useful they find this exercise.

Posted by Georgina Spencer @ Keele Pathfinder Team |

Additional 'Early Adopters' for Keele Pathfinder

February 25th

At the end of January, staff at Keele's Faculty of Humanities Learning and Teaching Development Unit were provided with a respository login, workflow, and collection "space" to enable them to place resources in the repository. Some content has been added since then and some feedback will be sought, and an assessment made, of the whole process in the coming weeks.

The university archivies department is also showing an interest in adding digitisations of its most requested items to the repository, and Michael and Scott will be meeting with archives staff this week to discuss workflows, metadata requirements and potential use.

An update meeting is also planned for this week with the LARC project members who had been provided with a repository log in and workflow at the end of January.

Keywords: intralibrary, metadata, users, workflows

Posted by Scott McGowan @ Keele Pathfinder Team

Intralibrary User Conference

February 22nd 2008

Some of the pathfinder team have been in Edinburgh this week to attend the intraLibrary user conference.

We've come here to discover what new features are coming in the next release and to see what solutions are available for our current challenges.

The new product does seem to have extended the API's for remote access over the web services protocol. This should help us to make the links between the VLE and the repository more intuitive.

We found some good news on the scalability front as we have found that the repository can use a transparent http accellerator which means we can distribute the retrieval of documents over as many servers as we want. That should be enough to allow us to use the repository for video downloads.

We'll try that out and see whether we can get a feel for the size of the servers we need to use to accomodate the videos our health faculty want to distribute.

Posted by Jonathan Knight @ Keele Pathfinder Team |

Workflows, groups, users and collections

January 23rd 2008

Both Scott McGowan and myself had been finding the relationship between the various components of the repository somewhat confusing, and decided we needed a session to work through this. The practical focus of the session was twofold. Firstly to set up a group for the LARC project and tie this in with a workflow in order that materials could be added to the repository by the members of the LARC gerontology project, and secondly to ring-fence the CLA content so that it wasn't externally searchable or browsable by non-approved users.

To achieve the first outcome we adapted a workflow currently in use for the uploading of past exam papers which would allow the gerontology staff members to upload materials and fill in rudimentary metadata. The metadata fields these contributors would see was fixed by creating a metadata subset and attaching this to the workflow. We then tied the LARC workflow to the LARC group and settled on the membership of this group. After a few false starts we finally managed to tie all the roles, workflows and users together correctly and all appears to be functioning as required.

The knowledge we'd gleaned from working this example through was useful in our second task. We knew the ring-fencing process was achieved through assigning the materials to a collection, and such a collection already existed. However, it didn't appear to be functioning as we desired, and materials were still searchable through the web-ct interface and browsable by repository users outside of the CLA group. By combining another new workflow with the user group and the collection we worked out how to correctly orient the CLA material upon upload. This left the issue of the materials already uploaded, and the worry was that we may have to reload them into this new workflow from scratch, thus losing the current URL links. However, upon brief examination we found that we could shift items already in the repository into a collection fairly easily.

So we're now happily au fait with the use of, and relationships between, workflows, users, groups, metadata subsets and collections. Now onto the next task of incorporating the bulk upload function into a workflow......

Keywords: Metadata, workflows

Posted by Michael @ Keele Pathfinder Team |

Plans for continued funding...

It's budget bidding time and we need to bid for some money to move this pilot project into a production service. So as we're piloting a repository we need a system capable of delivering the content to the students. The problem is that we're running a pilot so we've got early adopters using the system who are keen and have lots of data. So we don't really know how that will scale to a general purpose system. We also won't have the production release of the software until February so we don't know what kind of load it will put on the server. Obviously by the end of the project we will have a much better idea of what we need to run the production system but we need to bid for the budget now.

We summarised our current knowledge as:

Size of database: Big
Network bandwidth required: Lots
CPU throughput: Fast
Disk throughput: Big, Fast and lots of flashing lights

and of course we need two of them to give us high availability in the face of a hardware or software failure.

We stuck our finger in the air and hung some seaweed out the window and came up with a number that covers as many contingencies as possible without it looking like we trying to buy a small island in the Caribbean.

Keywords: budget disks network

Posted by Jonathan Knight @ Keele Pathfinder Team |

November Meeting with Intralibrary

November 29th 2007
The Keele Pathfinder team held a teleconference with Sarah Currier from Intralibrary on November 29th to review the company’s recent development work on the repository client software.

This work focused on the requirements to store, catalogue and extract metadata for the resources placed in the repository digitised under the Copyright Licensing Agency’s (CLA) Trial Blanked licence. The extraction of data is particularly important, as this relates to the potential use of the software to generate the reports required for the CLA. Sarah provided a remote demonstration of the metadata template developed as a result of recent discussions with the Pathfinder team, as well as a tour of the prototype CLA reporting functions. The Keele team were very pleased with the work that had been done, and Intralibrary will now take this forward into the v.3.0 Beta build of the repository software. The metadata template for CLA materials will include a field for course description, and will also include course “presentation” (i.e. a code that identifies exactly when a course is run).

Sarah usefully mentioned the XCRI (eXchanging Course Related Information) JISC project.
This project is working towards development of a standard specification (the “XCRI Course Advertising Profile”) using Extensible Markup Language (XML) to describe courses, and also to allow for this information to be aggregated and searchable. At the moment there’s no standard way to exchange information about university courses, but this project has a number of participants in HE working towards creation of such a standard. A detailed overview of the project is available from http://www.xcri.org/download/overview.pdf. The XCRI project is apparently interested in the Keele Pathfinder group’s activity, and it’s hoped that since Intralibrary is an XML based application, some dovetailing with this developing standard may be possible.

The Keele team anticipate the inclusion of these developments in the next build of the repository software, which is anticipated hopefully just before the end of 2007.

Keywords: CLA, Course, Keele, Metadata, Module, Pathfinder, Template, XML

Posted by Scott McGowan @ Keele Pathfinder Team