Intute Integration
The evolution of the Web has changed the way that people access information. Web 2.0 technologies have allowed information providers to integrate their services in people's existing online spaces, and users expect to be able to synthesise, edit and customise content for their own specific purposes. Intute, the JISC-funded service that aims to offer the best of the Web for Higher and Further Education, has responded to these changes by developing a variety of integration services which offer flexible ways of delivering its content to users. This article aims to give an overview of the Intute Integration project, which began in March 2006. It looks at the rationale behind the project, the integration methods offered, technical issues, a case study, uptake, challenges presented, and future plans.
Intute is a JISC-funded service that offers a searchable and browsable database of Web resources, all of which have been selected, evaluated and described by subject specialists working within the UK's Higher and Further Education communities. It also offers the Virtual Training Suite and Internet Detective to develop information literacy. Intute has always offered a range of content integration services, originally referred to as 'embedding' services, and many of these tools began their development in the Hubs that made up the Resource Discovery Network, Intute's predecessor. For example, the MyIntute service, which enables users to customise Intute content, was a direct descendent of the MyHumbul service from Humbul the Humanities hub. Further information about the Resource Discovery Network and its transition to Intute is available [1] [2]. The Integration Project represents a more focused approach, to develop and market them as one clear service, which supports Intute's strategic objectives and business plan.
Rationale
One of Intute's objectives is to demonstrate its efficiency and value to its funders, and offering integration services is one way of doing this. Given the current trend for re-purposing, sharing and mashing up content in various formats, Intute has developed flexible ways of offering free content to its user community. Intute is keen to extend its user base and one way of attracting users already bombarded with many online services is to offer access to Intute content in places where they already go, such as their own library portals and VLEs (Virtual Learning Environments). Intute also offers a personalisation service (MyIntute) which could appeal to many current users and which embodies many Web 2.0 functions for mashing up and redistributing content.
Integrating can save HE and FE institutions from 're-inventing the wheel' and individually producing similar guides to the Web. There can be cost savings for them too. Leeds University is a partner in an Integration Project case study with Intute and has saved £22,000 on replacement costs for hardware and software. Intute is committed to promoting this approach with universities and colleges and one of its Key Performance Indicators relates to this activity.
With the constant speed of change in the information landscape, Intute needs to keep abreast of developments in educational technology and offer progressive services. It also needs to offer flexible options which will appeal to users. According to the CIBER report, Information Behaviour of the Researcher of the Future [3], users show 'clear differences in information-seeking behaviour by subject, gender, and by work role'. The report investigated new types of online behaviour, which is of great interest to Intute. Many people are 'skimming' resources quickly, using only a fraction of content in a service, or relying on favoured brands. Users appear to want simplicity of use without advanced search mechanisms. Intute follows Web 2.0 developments and ideas, and aligns its objectives with JISC strategies, even if the informative What is 2.0? report [4] from JISC admits, 'people still can't actually agree what it means.'!
Ways of Integrating
Various methods of integrating content are offered, which can be broken down into five main areas:
- Linking to Intute or the Virtual Training Suite
- Search options, eg. Federated search, integrating the Intute search box
- Using Intute newsfeeds (over 100 available)
- Using MyIntute to customise Intute and non-Intute resources
- Users contributing content to Intute
More information on these methods is given on the Integration page of the Intute Web site. [5]
There are some interesting examples of universities or organisations which have integrated Intute content already. It can vary considerably. Most UK universities (66% in our August 2007 survey) had only used the simple link facility so far. Some had been more ambitious and put in a search box, newsfeeds and pulled in content from MyIntute. Bookmarks of more advanced integration examples can be viewed on del.icio.us. [6].
Technical Approaches and Issues
Intute offers two different technical approaches to providing integration options for our users. The first involves the provision of standard APIs (Application Programming Interfaces) that can be used by Web application developers to incorporate searches of the Intute catalogue into their own services. Library portals are good examples of users of this kind of service. The second approach provides services aimed at Web site administrators, often information professionals, who are not programmers but who wish to insert content from Intute directly into their own Web pages. Examples of this type of service might be the insertion of RSS feeds provided by Intute directly into user Web pages or the Intute Include service, which places an Intute search box into a Web site.
Intute APIs
Z39.50 and SRU/W
Despite its age, Z39.50 is still in demand as a programming interface to Intute's catalogue, generating just under 100,000 hits per month on average, the majority of which come from library portals such as Metalib. We chose a convenient Z39.50 interface that could be customised to work directly with our catalogue's database and which also provided an SRU/W interface out of the box. We had some initial problems with the unexpected demand via this link and had to restrict session times to prevent our server being overwhelmed by too many simultaneous user sessions. The software is highly configurable and we have expanded our SRU/W output from a basic MARCXML format, to a more standards-compliant version of MARC (MARC21slim) plus Dublin Core. XML output from SRU/W has also occasioned the usual headaches with odd character sets within the database, but all should now be output in the correct utf-8 format.
OpenSearch
OpenSearch [7] is a method for sharing the output of search engines by providing a means of specifying to an application how to submit searches to a search engine via URL and standard ways of outputting the search results using XML. Results can then be integrated into aggregating services or incorporated into the toolbar search boxes of Internet Explorer 7 or Firefox 2.
(There are simple instructions on how to do this [8]).
OAI-PMH
This protocol for metadata harvesting has long been used by the RDN, and now Intute, for exchanging metadata with partner services and others. In particular it is used with Google to ensure it indexes all our records, an important source of access to the catalogue. It is also used with OAIster, which indexes those of our records that point to full text documents [9].
Integration Services
For those who want Intute content within their Web site but do not have the desire or facilities to make use of our programming interfaces, we provide some applications for integrating our content.
Intute Include
Perhaps our main method for making content available via our users' Web sites is Intute Include. This is a mechanism whereby an Intute search box can be inserted into someone else's search page, with the search results formatted using that Web site's own look and feel. This is achieved by inserting the search results from Intute into a locally created template. This gives the impression to the site's users that they are using a service provided by the site, without navigating away to Intute.
The model used is based on a system developed at UKOLN for the RDN [10], the predecessor of Intute. In fact, when Intute was launched, our interim solution was to copy across the RDN-Include code, and alter it to work with the new Intute search engine.
However, the RDN-Include code was by this time quite old, returning non-compliant fragments of HTML, using tables for formatting, for insertion into the Web site's template. The code has since been rewritten using the OpenSearch standard as the basis for exporting search results (see above). Some extra functionality was also added so that full record details can be viewed, with keywords hyperlinked for further searches. Intute Include requires a Web administrator to install some simple software on the local Web server. This is only available for Linux/Unix systems at present, but we will be making a .NET version available within the next few months for Microsoft Web servers.
As some technical effort is required to install Intute Include, which may deter some users, we also make available Intute-Lite, which only requires the creation of an HTML page template for results and some HTML entered into the user's Web page. We suspect a lot of people avoid this option too, thinking that JavaScript is required, but a non-JavaScript version is also available to avoid any accessibility issues.
The main issue we still have with Intute Include is that is does not work with Web sites produced through the use of content management systems; so we will be looking at other models for embedding our search boxes into Web pages in the future.
RSS
Based on RSSxpress, also code originally developed by UKOLN, Intute-Channel-Lite enables RSS feeds (including Intute's 100+ RSS feeds, of course) to be directly embedded in a Web page without any technical know-how. This does require the use of JavaScript, however, which fetches the RSS feed via some software running at Intute that formats the feed as HTML, and inserts it in the Web page. The HTML uses CSS classes, which allow the user to format the output to fit with their own look and feel.
MyIntute
The MyIntute service is the user's own personal workspace, which allows users to save, tag and customise Intute content. Non-Intute Web sites can also be saved, tagged and exported. Exports can be done using JavaScript which enables dynamic updating of Intute records. Tagged records can also be emailed to yourself or other people, and saved Intute searches can be exported as RSS feeds. Further developments are underway, for example, allowing users to comment on and rate Intute resources, similar to the Amazon Web site.
Figure 1: Example of MyIntute screen with a tag cloud [11]
http://www.intute.ac.uk/myintute/index.php
The MyIntute system uses a combination of PHP, Perl, JavaScript, CSS and a MySQL database backend. Data from the system can be exported by the user as ordinary HTML, JavaScript, plain text, or, experimentally at the moment, RSS. The JavaScript and RSS exports give the user a pointer to a file which is saved on the MyIntute server. In this way, these exports can be maintained by the system and changed records can be updated without users having to do anything themselves. These updates are carried out each night when a scheduled script is run. Users also have the option to refresh the content at any time by re-exporting.
Some of the problems confronted when building the MyIntute system included the way different browsers render the CSS, as well as some small differences in JavaScript. The CSS in particular was designed to be as platform- and browser- independent as possible, though a couple of work-arounds still had to be used to account for differences between Internet Explorer and Firefox browsers.
MyIntute is currently being assessed for its accessibility, and a number of improvements have already been made. Some of the MyIntute functions rely on the use of JavaScript in the user's browser. Although all the major functions can now be accessed without JavaScript, there are still one or two which will not work for the non-JavaScript-enabled browser. This issue is being addressed.
The future development of MyIntute should see much more extensive use being made of RSS as an alternative to JavaScript for dynamic exports. It is also intended to start making use of the social networking possibilities of having registered users; for instance, by putting users with similar subject profiles in touch with one another. Leveraging the existing social networking tools, such as Facebook, may be one way to achieve this objective and a prototype Facebook application is already in development for MyIntute.
Marketing and Publicity
With the technical developments complete, the focus turned to promoting the Integration services to potential users. The main target audience was academic librarians. Intute consists of four subject groups (Arts and Humanities, Social Sciences, Health and Life Sciences and Science, Engineering and Technology) and one or two representatives from each group were invited to join an Integration planning group. A workshop was held in June 2007 to plan the first publicity campaign. The intention was to exploit Intute's advantage of having many contacts right across their subject communities. It was also decided to produce a series of Quick Guides, four in total, describing how to use the available services.
The first stage was an email campaign, aimed at pre-existing contacts in each Intute subject group. A standard email was sent out to contacts, either associations or individuals, explaining the benefits of Integration and pointing users to the Quick Guides. Then press releases were sent to key organisations or journals such as JISC, CILIP and Information World Review. News items were also posted on the Intute blogs. A standard Powerpoint presentation on Integration was written for Intute staff to adapt and use at publicity events.
To help users get started, we produced a DIY package of help materials in Summer 2007. The Integration help pages were updated, four Quick Guides [12] were written and the Intute online helpdesk was made available to take any email queries.
These guides are aimed at staff with a good knowledge of editing Web pages, although they could be a taster for less technical people who are curious. They form an introduction to Intute and the methods above, to explain more and show some pre-existing examples as screenshots, from universities who have already integrated. The guides were written with user in mind and tested by a non-technical member of the team, who set up a Web site. The Quick Guides were trialled on volunteers in the user community, which was a valuable exercise and provided some feedback and ideas for improvements.
The second stage of the publicity campaign is more focused, aimed at Intute's partner universities where staff work for Intute and at UK universities with the largest student populations. Intute staff are currently meeting library staff or e-learning managers at target universities to present on Intute Integration and discuss their needs. Staff also include Integration in other presentations on Intute and some do live demonstrations of MyIntute, to show its capabilities of exporting content in flexible ways.
The latest development is a new "Exemplar" programme. Institutions who have systematically integrated Intute into their Web sites are eligible to become Intute Exemplars. Further information is available [13].
Leeds Case Study
In the mid-1990s, the University of Leeds adopted the Roads software to list subject databases and Web sites on its library Web pages. This was in the pre Google era when Yahoo was a directory service and subject gateways were in their infancy. By 2006, this was creaking at the seams: the server needed replacing, the software was no longer supported and the database indexes were full. In addition, there were workload issues of maintenance and the Library wanted to avoid flat Web pages of links. They therefore approached us for advice.
A pilot was set up to test the feasibility of using Intute's new integration services to populate the Subject Guides to Internet Resources. A number of subject librarians agreed to be guinea pigs: Education, Law, Engineering, Dentistry and East Asian Studies. The pilot focused on workload, availability of Intute records and technical issues.
It explored a number of ways of adding Intute services into the library Web pages and of including content from the Intute catalogue:
- Embedding the Intute Search Box
- RSS feeds
- Customising resources using MyIntute
The result was a subject guide entirely populated by Intute and updated automatically:
Figure 2: Leeds University Library subject guide [14]
http://www.leeds.ac.uk/library/subjects/
After the trial, a training session was held for all subject librarians in December 2006 and by summer 2007 Intute's integration services were in use across the Web site.
In terms of workload, the librarians found MyIntute very quick and easy to use. For the majority of subjects, Intute had comprehensive coverage and in some cases better Web sites were found. There were some issues of scope and language – some of the Web sites which they wanted to include fell outside the scope of Intute's Collection Policy e.g. local NHS Trusts, while some Web sites were in languages not covered. The solution was to develop additional functionality which allowed users to add their own records in their own MyIntute workspace. In fact the only issue was one of Accessibility. To present the lists, MyIntute uses JavaScript which was not permitted on the Leeds University Web site. Local IT staff developed some PHP coding which solved the problem. Intute is now looking at this approach and working on other options, e.g. RSS feeds.
The project realised benefits for both organisations. Leeds University made real savings in software and hardware as well as in staff time on link and content checking. 'MyIntute has offered us an easy and effective way of promoting quality assured websites to our students', remarked Martin Gill, Faculty Team Leader, University of Leeds Library. At the same time, Intute received invaluable feedback on its products during the development stage and has increased access to its resources beyond the Intute Web site.
Measuring Uptake
Desk research was conducted during Summer 2007 to determine a benchmark of the impact of Integration services. The Universities UK listing was used to find university Web sites [15].
University library Web pages were browsed and searched for any examples of Intute content integration. Incidental outdated RDN links were also noted and the relevant libraries contacted to request an update of links. Results were recorded on an Excel spreadsheet. Unfortunately it was impossible to find out how many institutions had integrated Intute content in direct response to our publicity email. Some library staff may have heard of Intute from other sources or simply came across it by chance. The most popular method of integrating was a simple link to Intute - 66% of the survey population. Only 7% had implemented the Intute search box. 18.5% still had outdated RDN links – sometimes next to Intute links. Typically there were two categories where Intute was mentioned:
- Under a general heading like 'Electronic Resources' or 'Databases', along with other free and subscription services. Certain libraries sub-divided Electronic Resources further and put Intute under 'Gateways'.
- In addition, Intute was also often listed in 'Subject Guides' or 'Subject Resources'.
Clearly there is still scope for further integration of Intute. More research is needed to find out why libraries are mainly using simple links and how they would like to use Intute content in future. This survey will be repeated at regular intervals.
Below are some sample screenshots of integration examples:
Figure 3: Intute at University of the West of England [16]
http://www.uwe.ac.uk/library/resources/general/gateways/
Figure 4: Economics Network use of MyIntute to import economics records [17]
http://www.economicsnetwork.ac.uk/links/general.htm
Conclusion
There is great potential for further integration of Intute content. Intute needs to do more user research to find out how and why integration is being carried out (or, indeed, why not). As Intute researcher Paul Ayres succinctly puts it in his article for ALISS Quarterly [18], 'The next part of the challenge is surfacing Intute data in truly open parts of the Web, which means engaging with and contributing to third party websites.' Integration needs to be made even simpler if possible, to encourage more users to make use of our content.
References
- Retrospective on the RDN, Debra Hiom, April 2006, Ariadne, Issue 47 http://www.ariadne.ac.uk/issue47/hiom/
- Intute: The New Best of the Web, Caroline Williams, July 2006, Ariadne, Issue 48 http://www.ariadne.ac.uk/issue48/williams/
- Information Behaviour of the Researcher of the Future (CIBER) 2008 http://www.jisc.ac.uk/media/documents/programmes/reppres/gg_final_keynote_11012008.pdf
- Anderson, Paul: What is Web 2.0? JISC, 2007 http://www.jisc.ac.uk/publications/publications/twweb2.aspx
- Intute Web site Integration page http://www.intute.ac.uk/integration/
- Del.icio.us bookmarks http://del.icio.us/angejoyce
- OpenSearch http://www.opensearch.org/Home
- OpenSearch instructions http://www.intute.ac.uk/integration/opensearch/
- OAIster http://www.oaister.org/
- RDN-Include: Re-branding Remote Resources: A Poster Accepted for the WWW 10 Conference http://www.ukoln.ac.uk/web-focus/papers/www10/
- Example of MyIntute screen with a tag cloud http://www.intute.ac.uk/myintute/index.php
- Intute Quick Guides http://www.intute.ac.uk/integrationguides/
- Exemplar Quick Guide http://www.intute.ac.uk/integration/exemplars/
- Leeds University Library subject guide http://www.leeds.ac.uk/library/subjects/
- Universities UK List of University Heads http://www.universitiesuk.ac.uk/members/
- Intute at University of the West of England http://www.uwe.ac.uk/library/resources/general/gateways/
- Economics Network use of MyIntute http://www.economicsnetwork.ac.uk/links/general.htm
- 'Is the Web2.0 Price worth paying? Learning from Economics',Paul Ayres, ALISS Quarterly, v. 3, no.2, (pp. 18-21), 2008. ISSN: 1747-9258