Making OER visible and findable

CETIS has recently updated our guidance to OER Projects. To supplement this, we’ve been thinking about how to ensure that your content is as findable as possible. There are many routes to finding the content released by UK OER projects, including through Jorum, Xpert and search engines. Myself and David Kernohan have drafted the following guidance to help projects think through the best ways to make OER visible and findable.

Who do you want to be able to find your resources? How can you help them do this?

Firstly, know what RSS functionality your platform offers. One lesson that came clearly from the phase one OER projects was that you choose a platform for all sorts of reasons, it’s rarely a choice from a blank sheet of paper. Whether you are using iTunesU or an institutional e-prints repository or a homegrown solution, make sure you know how it exposes its content to the web so that you can optimise it for discovery. RSS feeds need to be easily findable, both by people and machines, and contain enough information about the items to be useful in identifying the items.

Considerations:

1. What sorts of rss feed functionality does your chosen platform offer? Do you understand the options?

2. Have you got an rss reader set up to display the feeds, so that you understand how they work?

3. Do you have an rss logo clearly visible from your homepage, does it indicate there is a feed of OER content available (whether from your own website or elsewhere)?

Imagine someone has heard of your project. Via mailing list, a casual comment from a colleague, a tip from a tutor, or your own publicity. The first thing they would do is head for your website, maybe via a query to a search engine if they don’t have your url.

4. Is your content findable using major search engines?

5. If your content is hosted outside your institution, what comes up first in the search results, the content or the project website? Do you want to optimise both for search engines, or choose one as your main presence?

6. Is your project  website  findable from the home page of the institution or network you are linked to? (by browsing and/or by searching)?

If they find your project website, they would immediately ignore all of your fine project documentation, your “about the team” list and your contact details. They would head straight for the resources. so it matters how your project website links to your content (wherever the content is held).

7. Are your open educational resources visible from your project website? Is there an obvious link?

8. When you get to the resources (either from the index of your website or a specific “resources” page) do they look attractive? Is it easy to browse, search and identify resources?

So, someone has found the resource, whether on your own site or a third party site. On finding a resource that interests them, our user may want to view, bookmark or download the resource for later use.

9. Is it possible to view your resource (or at least enough information about the resource to allow people to understand what it is) from your website?

10. Is it possible to directly link to a resource for bookmarking or sharing?

11. Is it obvious how someone can download a resource? Is it obvious what someone needs to do with a resource when they have downloaded it?

The use of the downloaded or direct-linked resource may happen some time after it’s discovery. At the point of use, our user would need to know how the resource can be used, and what conditions would need to be fulfilled in order to meet the requirements of the license.

12. Is license information available within the actual resource, now it is downloaded?

13. If a license requiring attribution has been used, is full attribution information available within the resource?

We are increasingly seeing the rise of OER aggregators, automated tools that collect information about OER wherever it is stored and allow people to search across it. Aggregators have the potential to become a key means of discovery for OER. Examples of aggregators include:

Different aggregators have different requirements and it is worth properly examining these needs before you attempt to have your material included on an aggregator. CETIS is working on some advice about registering to aggregators. There are some needs that all aggregators have in common. First off all, an aggregator will need an RSS feed of your material (or a choice of feeds).   Like an RSS feed for a blog or a news site, each “item” has its own address that leads directly to that item. The feed also describes the item, allowing aggregators and people to know what the item is and how it should be described to others.

Machine readable feeds are important, and its also important that humans find your OER entry pages attractive. We’d also like to have a showcase like the  JISC content portal, for very little money we could create one for UK OER content, but for it to work we need good attractive homepages describing the nature of the content you have released and linking clearly to the content, wherever it lives.

We think the wealth of content created so far would benefit from a little extra push to make sure it is found in all the right places.

Post by Amber Thomas and David Kernohan. With particular thanks to Phil Barker’s recent post on sharing service information, Pat Lockley’s tireless crusade for better quality RSS to feed Xpert , and colleagues across the OER projects.

Comments

14 Responses to “Making OER visible and findable”

  1. Pat on December 6th, 2010 5:01 pm

    If any one wants any RSS help feel free to get in touch.

  2. Amber Thomas on December 6th, 2010 8:08 pm

    An additional point from Phil Barker is that best practice would be a “technical about page” describing the feeds that are available, where the OAI-PMH base is and what the API does. He’s quite right, and this is also something JISC would like to strongly encourage.

  3. Andy Powell on December 7th, 2010 11:27 am

    Re: points 4 and 5 (about search engines)… I suggest that people also think about what search terms they might reasonably expect people be using in search engines to find their content. Then work on making it as highly-ranked as possible against those terms.

    High ranking typically requires inbound links (as many as possible) – it might be sensible to think about how such links can be encouraged at the programme level. Pushing out RSS feeds (with links back to individual items) is certainly one way.

  4. Phil Barker on December 7th, 2010 11:47 am

    I think these are really important issues being discussed here. I especially like the “read your own RSS feed” line. That generalises well to “use your own API” and “harvest through your own OAI-PMH provider” etc…the point being if your machine interface is broken, how would you know?

    I would add as a complement to the “Imagine someone has heard of your project.” line in this post, “imagine if someone hasn’t heard of your project but by looking for a resource they need find one of yours”. Their entry point to your site will likely be the resource (imagine if that’s a pdf or word doc) what can they find out about the licensing terms for the resource, other related resources, your project, your institution with that as a starting point?

    Also, as well as advertising your RSS feed with a logo on the page it wouldn’t go amiss to use the link element in the header (e.g.
    ) which is what some browsers use to put the RSS logo in the address bar.

  5. Phil Barker on December 7th, 2010 11:48 am

    The example for the link element got stripped from my comment. Here it is with square instead of angle brackets
    [link rel="alternate" type="application/rss+xml" title="RSS 2.0" href="http://infteam.jiscinvolve.org/wp/feed/" /]

  6. Nick Sheppard on December 7th, 2010 11:57 am

    With respect to the emphasis on RSS…due to idiosyncrasies of our repository platform at Leeds Met, our RSS feed were not suitable either for bulk import to Jorum or harvest to Xpert and we worked with Jorum in order to bulk upload IMSCP and with Xpert to harvest OAI-PMH.

    At the time, transfer of full IMSCP to Jorum was considered an ideal solution as the national service is keen to host actual content rather than just links (like Xpert), however, I’m now not so sure as it means that our resources are duplicated in Jorum which has an impact on OER tracking (and exposure of our local repository content as Jorum inevitably appears higher in Google results due to greater pagerank) and in some ways I think it would be preferable for Jorum records to point at our local install – OAI-PMH harvested records in Xpert *do* point at our locally curated files.

    I’m also interested in the possibility of building value added services on centralised repositories of harvested data and we are currently working with Xpert for our ACErep project – http://acerep.wordpress.com/ – which has has a number of APIs available to search their data (thanks Pat!).

    Also, I think OAI-PMH is rather richer than RSS in terms of metadata harvest so may offer the possibility of building more sophisticated bespoke search portals for specific groups of stakeholders (e.g. medical) than can be achieved by Jorum.

    I’ve blogged about some of this:

    http://repositorynews.wordpress.com/2010/11/26/repository-reports-and-more-on-seo/

    http://repositorynews.wordpress.com/2010/09/17/oer-repositories-and-preservation-the-elephant-not-in-the-room/

    http://repositorynews.wordpress.com/2010/10/01/xpert-vs-jorum/

    http://acerep.wordpress.com/2010/11/23/working-search-prototype-and-a-sword-fight/

    In addition to the now famous Xpert, ukoer colleagues may also be interested in Ariadne Globe – a European project based in Belgium (I think) – http://ariadne.cs.kuleuven.be/GlobeFinderF1/search/ – which is another OAI-PMH harvester/search portal that has “a REST interface over HTTP with a query in JSON and result is in JSON. It is planned to have also plain parameters in the GET querystring and responses in XML. In that way anyone can build there own browsing environment if they wish.”

  7. Tracey Madden on December 7th, 2010 12:13 pm

    Think it very important to return to the question of ‘How do people look for/find resources?’. When we know this we can take resources to them. It’s not feasible to expect them to trawl through all the possible sites where there may be OER and (sorry to state an obvious but important point) they are not going to go to sites they don’t know about.
    This is not just true of OER this is true of all resources and project outputs that just don’t get noticed as widely as they should and hence don’t get used as much as they might.

  8. Lorna M. Campbell on December 7th, 2010 12:33 pm

    @Phil Agree about users coming to project websites from resources rather than vice versa. Easy to imagine a use case where a user finding a resource through a search engine or harvester might be interested in other resources from the same collection.

    @Andy better understanding of search terms is an important point. We discussed this at the #cetiswmd meeting and are hoping to do some work in the area of search log analysis.

  9. dkernohan on December 7th, 2010 1:46 pm

    @Tracy that’s a very good point, thank you for making it.

    But if we are going to host resources outside of Jorum too (and I still think that this is a good idea for so many reasons) then it makes sense to make them attractively findable.

  10. Nick Sheppard on December 7th, 2010 2:49 pm

    If Jorum aggregated, say, our OAI-PMH and we were able to get at that data via an open API, our OERs would be discoverable via Jorum AND we could also build (a) bespoke portal(s) using said open API…Jorum doesn’t have an open API though which is why we needed to look to Xpert instead:

    http://repositorynews.wordpress.com/2010/10/01/xpert-vs-jorum/

  11. Stuart Dempster on December 7th, 2010 3:32 pm

    A number, though not all, of these issues have been addressed during the Netskills short courses on optimising your web presence funded by JISC during 2010. A “guide to good practice” is scheduled for release in Q1 2011. We will endeavour to address some of the above in it.

  12. Alastair Dunning on December 8th, 2010 11:12 am

    Another key point in all this is the importance of having stable and accessible URIs. URIs that can be easily shared by humans, but also by machines, are the bedrock of good exposure. For OERs, creators need to devise a sophisticated URI policy that allows for collections of resources to be easily found and exposed, but also the individual items within a resource. In terms of reuse by other institutions, it is the ability to get hold and reuse elements of a resource that is probably the most important.

    There is also the issue of how technical exposure of content maps into a broader dissemination strategy. To enable plenty of inward links (and therefore help Google ranking) requires content creators to build up contacts with other interested parties who will recommend and support the OERs. If learned societies weren’t quite so stuck in the twentieth century, they would have a important role to play in this, recommending and linking to OERs within their specific subject areas.

  13. Making OER Visible | on December 10th, 2010 12:21 am

    [...] OER Visible Posted on December 10, 2010 by openedblogger| Leave a comment Amber Thomas has a new post about making open educational resources “visible and findable”. From the [...]

  14. OER Visualisation Project: Adding a bit of OAI-PMH collecting activity data in Google Spreadsheets [day 5] #ukoer #ooher – MASHe on December 9th, 2011 4:16 pm

    [...] of open resources but hidden away. Here are some tips from the JISC Digital Infrastructure Team on Making OER visible and findable.Share this:window.___gcfg = {lang: "en-GB"}; (function() { var po = [...]

Leave a Reply