SIVITER, Douglas, Dr.
School of Computing, Information Systems and Mathematics, South Bank University, London SE1 0AA, UK. D.Siviter@sbu.ac.uk http://www.sbu.ac.uk/siviter
Abstract: 'DisCourses' is the name of a newly formed collaborative educational network that aspires to support learners, teachers and facilitators of courses in the subject area of Developing Information Systems. The primary aims of the DisCourses network are to enable participants from a range of educational institutions to exploit informal collaborations, exchange ideas, share good educational practices and to technically share distributed resources and services. A similar collaborative educational network called 'NooLearning' has equivalent aims in the subject area of Object Orientation. This paper describes some of the techniques and tools used to support the collaborative shared provision of educational resources and in particular it focuses on the role of resource libraries in pursuing those aims.
Keywords: Collaborative Educational Networks; Developing Information Systems; Object Orientation, Resource Gateways; Web-based Educational Resources; Lotus DominoTM
The DisCourses network [1] and the NooLearning network [2] started as collaborations between the computing departments from four UK Universities. These networks are attempting to generate user groups of participants from a range of institutions, focussing initially on UK educational institutions but expanding where feasible to any interested parties world-wide. The projects to develop the DisCourses and NooLearning networks are continuations of a previous UK Universities collaboration [3] that developed interactive courseware resources for software systems engineering.
One of the aims of these networks is to enable the collaborative shared provision of distributed educational resources. This is regarded as a recurring aim within many subject disciplines. Hopefully a description of some of the issues and technical challenges that arise when pursuing this aim will be of interest to people working in any discipline.
Specifically, this paper describes:-This discussion of collaborative shared provision is concerned with current pragmatic approaches to 'picking and mixing' from existing distributed educational resources in order to provide new combinations of educational resources that serve local requirements. The discussion is not about the long awaited next generation of highly interoperable courseware objects [7] that offer seamless plug and play in a variety of educational contexts. Such advanced courseware objects built to accepted international standards and built from super software components are frequently described as belonging to that elusive 'next generation', always just out of reach, and not yet exploitable by today's academics. What current academics face instead is a world where low-tech, non-standard educational resources are proliferating, having been enthusiastically developed by hundreds of relatively isolated academics that successfully dabble with programming languages, multimedia authoring tools, and of course web publishing tools. Although these educational resources can be criticised as low-tech and in software interoperability terms thoroughly non-standard, they have the virtues of being currently available and they embody a substantial amount of educational investment and expertise.
Academics increasingly appreciate the opportunities offered by Computer Assisted Learning and the opportunities for distributing and accessing materials via the World Wide Web. One sign of this growing appreciation of opportunities is the emergence of collaborative educational networks within which attempts are being made to improve the efficiency with which people can share their educational products. There are clearly many barriers to overcome (both technical and cultural) before large-scale examples of virtual courses operating across virtual universities become ubiquitous. However, it is now totally realistic for experiments to be conducted in overcoming these cultural and technical barriers. These experimental and affordable opportunities are available to geographically dispersed groupings of academics in ways that were impossible prior to the emergence of the web.
So, a typical scenario, and one which is central to this paper, is the idea that lecturers working in various universities are now routinely capable of supporting their teaching by developing web-based resources. The lecturers all know that many other lecturers are also developing similar resources and they are sufficiently mature to have overcome the 'not-invented-here' syndrome such that they are now keen to both share their own resources and to re-use other people's resources whenever practically feasible to do so. Readers will no doubt be amused by the number of contentious idealistic assumptions stated in the previous sentences but the ideals are still worth pursuing. They are the ideals being pursued by both the DisCourses and NooLearning networks. A key phrase was 'whenever practically feasible' and it is one of the aims of the networks to develop techniques and tools that increase the feasibility of effective sharing of distributed educational resources. One key idea that emerges is the role of resource libraries to support collaborative educational networks.
Consider again the primary aims within the DisCourses and NooLearning networks, i.e., "to enable the shared provision of on-line educational resources", "to enable pragmatic approaches to 'picking and mixing' from existing distributed educational resources in order to provide new combinations of educational resources that serve local requirements". Seen from a lecturer's perspective this can be summarised as "how do I effectively search for educational resources on the web?", "how do I audition these resources?", "how do I plug them into the resources that I am developing for local use?", and "how do I reciprocate and offer my resources to the community?" The DisCourses and NooLearning resource libraries are being developed specifically to support these aims.
Consider first how a lecturer might go about searching for and auditioning educational resources on the web. It is informative to compare the resource libraries with other web-based facilities, in particular search engines and information gateways.
The classic general-purpose search engines available on the web (e.g., altavista.com, yahoo.com, hotbot.com, and many others) are immediately regarded as too broad and too general. Despite their tremendous utility in other contexts these search engines do not provide appropriate catalogues of specialist educational resources, and any attempt to search for, audition and then assemble and re-assemble educational resources using these tools is very laborious.
More appropriate facilities, e.g., information gateways, have emerged from research and developments in the world of digital libraries. Typically, an information gateway is a browseable and searchable repository of links to resources that have been nominated and vetted and described and catalogued by people acting in a subject specialist and/or librarian capacity. The hope that underpins an information gateway is that these human editorial activities will add significant value to the resources. These activities are supposed to raise the probability that resources discovered through an information gateway will be more relevant and more useable by people seeking those resources. A lecturer searching for resources on a specialist topic may prefer to retrieve ten highly relevant candidates from an information gateway rather than retrieve two million mostly irrelevant hits from a web search engine. Richly described resources (i.e., resources plus appropriate metadata) will be much easier to browse, search, filter, and 'read about'. The process of auditioning educational resources can be streamlined by auditioning the metadata.
Information gateways are being developed in many shapes and sizes and there is significant diversity in their scope. There are also laudable attempts being made to generate and comply with standards for cataloguing and standards for achieving interoperability between gateways. So, for example, designers of information gateways aspire to offer descriptions of resources where the descriptions (i.e., the metadata) conform to well defined standards. They also aspire to offer interoperable facilities whereby an enquirer issuing a query to one gateway may find that the gateway has intelligently passed the query on to several other gateways in a chain and has then accumulated a collection of results for the enquirer. For further reading about these issues see Pinakes [8] which provides a launch pad for several examples of information gateways and see within the UKOLN site at [4] for discussions of metadata standards (e.g., Dublin Core) and interoperability standards (e.g., Z39.50)
The quest for standards in this kind of resource management is not just confined to the world of digital libraries. Within the specific subject area of educational technology there are major initiatives that are attempting to formulate standards. Hence, attempts to specify standards for metadata can be found in the Instructional Management Systems project [5] and in the IEEE Learning Technology Standards Committee [9]. Observations about both of these can be found in [7].
It is not at all clear how, if ever, the standards being pursued within the digital libraries initiatives and the standards being pursued within the educational technology initiatives will eventually work in harmony. This means that any present-day collaborative educational network that wishes to implement a resource library is faced with research challenges rather than stable available standards. Within the DisCourses and NooLearning networks the decisions were made to follow the path being forged by the IMS project while keeping eyes on where other initiatives are leading. The attempts to formulate what might be called 'catalogues based on IMS compliant metadata' within the resource libraries are still regarded as experimental research. In this respect, the DisCourses and NooLearning resource libraries differ from several information gateways; they lean towards metadata standards emerging from the educational technology world and are not yet committed to interoperability standards emerging from the world of digital libraries. Otherwise, the resource libraries aspire to offer the same functionality as information gateways, i.e., they enable browsing and searching of nominated, vetted, richly described, and well categorised resources.
Assuming that a resource library helps lecturers to find and audition educational resources, what about the next challenges? "how do I plug these discovered resources into the educational resources that I am developing for local use?", and "how do I reciprocate and offer my own educational resources to the community?"
Standards by themselves will not provide the answers to the issues of providing pragmatic approaches to picking and mixing resources. Within the DisCourses and NooLearning networks one issue that was quickly identified as crucial, and yet apparently overlooked or mostly ignored by information gateways, is the idea of durable URLs. A URL is a Uniform Resource Locator, more commonly referred to as a web address. Durable URLs are an essential requirement for authors that wish to invest time in finding auditioning and exploiting distributed resources. Authors require the assurance that their own resources, built out of hybrid collections of links to other people's resources, will not turn into a mess of broken links and/or a maintenance nightmare if and when those resources have a change of address.
To illustrate the importance of this point I would like to refer the reader to what the DisCourses network calls a system leaflet. A system leaflet is just a simple A4 sized printable leaflet that describes some aspect of the inner workings of the DisCourses facilities. The technical documentation for the DisCourses facilities comprises a series of these on-line leaflets, each of which has its own web address (URL). One such leaflet is entitled "The Role of Resource Libraries within Collaborative Educational Networks". This leaflet describes the importance of durable URLs in more detail. So what is the URL for this particular leaflet? Where can the reader of this paper locate this leaflet? If I publish the actual URL in this paper and then later decide that for some reason I need to change the URL, how do I notify the reader of this paper that the URL has changed? How can I even begin to find out who is experiencing the broken link? This is the essence of a link maintenance nightmare. A solution is to register the resource in a resources library and allow the library to issue a unique and durable URL for that resource. The durable URL when accessed by the reader will request that the resource library automatically re-directs to the current actual URL for the resource. This indirection is an ancient idea but it is also tried and tested and is well worth exploiting in the context of a community-wide resource library. Of course some update will still need to occur when a URL changes but in this context the provider of the resource needs to inform only one place, the resource library, in order to enable all users to carry on using the resource without interruption.
Given that this conference paper is intended to be published on the web, it is possible to illustrate the point practically, as follows:-
Below are some examples of resources that are published via the DisCourses Resources Library. Some are very small resources like system leaflets, others are larger resources like a web site to support an undergraduate unit. In each case, the resource is presented as a link. In the first row is a user-friendly name that an author would present to the user. In the second row is the durable URL that authors would be encouraged to use (but not display) as the link. This durable URL is NOT the URL of the actual resource. The URL of the actual resource will be retrieved by the library at the time of accessing the durable URL. End-users are discouraged from using a link to the actual URL. Note that this is the opposite of the practice encouraged by search engines and many information gateways.
Examples:
| The Role of Resource Libraries within Collaborative Educational Networks - a DisCourses System leaflet |
| available at http://library.discourses.ac.uk/resources/1999/0405/0001/ |
| Designs for Resource Libraries within Collaborative Educational Networks - a DisCourses System leaflet |
| available at http://library.discourses.ac.uk/resources/1999/0312/0001/ |
| An Introduction to Prototyping - a taught topic within an undergraduate computing unit |
| available at http://library.discourses.ac.uk/resources/1999/0228/0001/ |
| Techniques of Systems Analysis and Design - a taught unit within an undergraduate computing course |
| available at http://library.discourses.ac.uk/resources/1999/0412/0002/ |
| ACE lab facilities - a guest access facility for exploring collaborative tools in education |
| available at http://library.discourses.ac.uk/resources/1999/0412/0001/ |
| Helping institutions to collaborate in the shared provision of on-line educational resources - this paper published at the ICEE99 conference |
| available at http://library.discourses.ac.uk/resources/1999/0515/0001/ |
Note the flexibility of these durable URLs. Consider the last example. At the time of writing this paper, I have no idea what the actual URL will eventually be for the paper as published by the conference organisers. I can however, safely publish the durable URL as above. At the time of writing this paper the durable URL will re-direct to an actual URL that points to somewhere on my own server. Later, when the conference organisers mount the paper on their own server I will notify the DisCourses library to update the actual URL. If a reader had made links to the durable URL as above they would not need to make any changes when the actual URL gets updated. Also, if for some reason at some future date the conference organisers cease to maintain their own URL I can always revert back to my own URL, notify the DisCourses library about the change and again allow end users the convenience of not being required to make any changes to their own links.
The above examples accessed within the web-based version of this publication provide an opportunity to briefly discuss the non-technical issues of copyright, and publishing on the web. For example, the copyright intention for some resources published via the DisCourses library (e.g., the Prototyping topic listed above) states that the resource may be freely used in educational contexts provided it is accessed via the registered URL published by DisCourses. The resource may NOT be copied onto or mounted on other servers, etc. In effect the author has stated that links to the resource are freely allowed but embedded copies of the resource are not freely allowed. Taking the same approach for the DisCourses System Leaflets means that they are freely available to anyone to access via the DisCourses published URLs but not available to copy without the author's consent. This means that in this case, I have also explicitly granted that consent to the ICEE'99 conference organisers so that they may publish those leaflets (as embedded copies) within the appendix of this paper. This technical distinction between links and embedded copies is a recurring issue that affects any publishing relationship between authors, publishers and consumers of resources. A vehicle like the DisCourses resource library can help to bring the legal issues into focus and help authors to make their publishing intentions explicit even if it fails to offer immediate solutions.
Within the DisCourses and NooLearning networks, hybrids of HTML (web pages) and Lotus DominoTM databases are used as an enabling technology at many levels including:-
In section 5 there was a reference to a system leaflet that described "Designs for Resource Libraries within Collaborative Educational Networks". This system leaflet presents a layered model of a resources library within which a top layer comprises multiple catalogues where each catalogue provides multiple views of resource collections. Each entry in a resource collection is a Lotus Notes document specifically designed to record metadata for a registered resource. Lotus Notes/Domino is a superb vehicle for recording collections of rich documents, for generating multiple views of those collections, for supporting secure distributed access to the data and for enabling access to all these facilities via web browsers for most people but also via Notes client applications for more specialised access features. Given the experimental nature of the metadata cataloguing within DisCourses and NooLearning it is very important to be working with robust distributed tools that offer the programmability and flexibility of Lotus Notes/Domino.
In contrast, the next layer in the resource library design is referred to as the 'Registered Resource References' layer. This layer implements the durable URLs. For this layer the complexities of Lotus Notes/Domino are deliberately avoided. This layer is mostly elementary HTML and is deliberately separated from the complexities of the higher level catalogue facilities.
After authors have exploited the catalogue with its multiple views and search facilities, they will have located durable URLs for resources that they wish to exploit within their own resources. The authors have valid reasons for working with a sophisticated Domino-based catalogue while they are constructing resources but the intended end-users have no need for this elaborate catalogue, these users simply use the durable links identified and exploited by the author.
These durable links and the mechanisms that enable them to locate and re-direct to the actual URLs are implemented as simple HTML and JavaScript. The only links between the Domino-based catalogues layer and the HTML based Registered Resource References (RRR) layer are administrative links. The production and maintenance of the separate RRR layer is a task that is supported by the programmable Domino environment but the end-use of the RRR layer is totally independent of Domino.
Because Lotus Domino provides groupware-oriented facilities Lotus Domino is exploited in several other contexts quite separately from the resource libraries. The DisCourses and NooLearning administration is spread across several universities and relies on on-line collaboration. The traditional discussion databases and document store databases that come as standard Notes templates are customised very slightly and used to support the distributed administration within the network. The idea of exploiting a pair of databases (one threaded discussion and one document store) as a loose teamroom arrangement seems to work well and is therefore used in a variety of other contexts too.
Typically this teamroom arrangement is accessed from within some other HTML / frames design so that the Domino-based teamrooms are just one facility linked within and accessed from some higher level resource. This hybrid arrangement is so flexible that it is possible for members of the network to provide teamroom services to other members of the network in a similar manner to providing less dynamic educational resources. An interactive group-oriented service becomes just one more resource to register in the resources library for others to share. The difference is that services like these typically require more attention and more maintenance and are less likely to be offered to or exploitable by anonymous users. For examples of providing this hybrid arrangement of HTML / frames coupled with linked Domino-based teamrooms see the ACE labs guest access facilities referred to in the examples in section 5. You can also see an example of exploiting these ideas in context by referring to the Techniques of Systems Analysis and Design taught unit also referred to in section 5, however access to the fifty or more student teamrooms in this example will be restricted.
The techniques of implementing HTML / frames and Domino hybrids are also available for network members to exploit. Techniques at this level tend to be encapsulated within documented templates that can be downloaded and re-used in whatever way authors decide.
To summarise, hybrids of HTML / frames and Lotus Domino databases serving as teamrooms are a flexible and powerful combination that is exploited at different levels including for inter-institutional consortium administration, for intra-institutional team administration of taught units, and for student teamrooms to support students collaborating on group assignments. The teamroom service is regarded as just one more pluggable facility that can be hosted by one institution and exploited by another institution much like a courseware resource but more group-oriented. Finally, facilities, services and many kinds of educationally useful resources are all registered in and accessed via a resource library which is itself a hybrid of HTML / Frames and Domino.
| 1. | DisCourses Collaborative Educational Network,
URL: http://discourses.ac.uk |
| 2. | NooLearning Collaborative Educational Network,
URL: http://noolearning.ac.uk |
| 3. | W.I.S.D.E.N. Courseware Resources for Software Systems Engineering,
URL: http://www.it.brighton.ac.uk/wisden |
| 4. | The UK Office for Library and Information Networking,
URL: http://www.ukoln.ac.uk/ and URL: http://www.ukoln.ac.uk/metadata |
| 5. | IMS - Instructional Management Systems,
URL: http://www.imsproject.org |
| 6. | Lotus DominoTM
URL: http://www.lotus.com |
| 7. | Siviter.D., Educational Objects: from courseware widgets to virtual universities, Educational Technology & Society 2(2) 1999, ISSN 1436-4522.
URL: http://ifets.gmd.de/periodical/vol_2_99/siviter_article.html |
| 8. | Pinakes - subject launch-pad to UK National Gateways,
URL: http://www.hw.ac.uk/libWWW/irn/pinakes/pinakes.html |
| 9. | IEEE Learning Technology Standards Committee, P1484,
URL: http://grouper.ieee.org/groups/ltsc/ |
The DisCourses System Leaflets referred to in section 5 of the paper are presented here as appendices. Each of the two leaflets comprises 2 pages ready for A4 printing. Each page is produced as a gif image.
See Figures 1 & 2. for the leaflet entitled "The Role of Resource Libraries within Collaborative Educational Networks".
See Figures 3 & 4. for the leaflet entitled "Designs for Resource Libraries within Collaborative Educational Networks".
Figure 1. System Leaflet - The Role of Resource Libraries - Page 1 of 2 pages
Figure 2. System Leaflet - The Role of Resource Libraries - Page 2 of 2 pages
Figure 3. System Leaflet - Designs for Resource Libraries - Page 1 of 2 pages
Figure 4. System Leaflet - Designs for Resource Libraries - Page 2 of 2 pages