9. Recommendations

1. JISC commission some small pieces of work looking at user requirements for personalisation in well defined parts of services. The work to cover a spread of JISC-funded and institutionally-funded services. The work would not involve developing interfaces or the technical infrastructure to support them (if these do not exist already) but might involve sketching, storyboarding and mock ups.
2. JISC consider funding a small number of projects to explore innovative and tightly defined uses of personalisation approaches, particularly APOD and APUA in the context of JISC services and institutional portals. Such projects could either examine the application of the open source tools identified in 5.2.2.2 and 5.2.3 or address the gaps identified there. All such projects should include a strong evaluation element.
3. JISC fund and evaluate a small number of pilot projects looking at sharing personal profiles. These should fall into two categories:

1. Sharing profiles between organisations in the UK academic community. These projects should look at the practicalities of sharing and getting profiles to work. The issues of permission and centralisation or distribution of profile information should be examined.
2. Preliminary and investigative projects between diverse organisations, looking at issues of user trust and potential benefits and problems from sharing profiles. Since we are suggesting an examination of which institutions users might trust with which part of their profile, a number of different and novel partners might be considered. We would suggest financial institutions, mobile phone service suppliers (not manufacturers), sporting organisations (both those where users have the role of players and that of spectators), solicitors and trade and student unions.

4. We recommend that all institutions, and particularly those who have a highly devolved information structure, make it a priority to ensure that their data and processes can supply the needs of a personalised service. It is desirable to implement a personalised service which allows teaching staff to access up to date student and course data, students to ensure that their personal data is accurate and administrative staff to access reliable information from an authoritative source. It is also one way to throw a spotlight on insecure, inadequate or duplicated processes.
5. Every effort should be made to capitalise on JISC's substantial investment in services and resources. Users may want to use resources through personalised institutional portals or through subject based services. Promotional activity would be better focussed at service providers, whether institutional or subject-based, encouraging them to use the indisputably valuable resources that comprise for example the RDN rather than attempting to promote an RDN brand to end users. This applies to JISC services in general. It is inevitable that every service and project will have its own Web site and it is desirable for end users to have multiple interfaces to publicly funded resources. However it is undesirable that competition between government funded services for end-user "hits" leads to restrictive practices and discourages sharing. Assessment measures and performance indicators which encourage such restrictive practices should be removed and replaced with measures which encourage and incentivise resource sharing, nationally and internationally.
6. It would be helpful if within JISC there were a common vocabulary for describing personalisation and its results. We recommend that a modified and updated version of the base definitions used in this report (see Appendix 11.1) be widely used by JISC services, bids, projects etc to define what they are doing and will do by way of personalisation. A short, compulsory survey of all active projects and services would be a good start.
7. For information resources such as RSS newsfeeds to be used interchangeably in a personalised environment such as a portal, there needs to be a consistency of approach to implementation (e.g. the use of the 'title', 'description', 'category' and 'ttl' elements). We recommend that JISC produce guidelines for implementation of each element (in effect an application profile of the RSS specification) for use by all JISC services and that adherence to these guidelines be made a condition of funding. These guidelines should be promulgated widely to enable third party suppliers of RSS feeds to adopt a common approach and to inform consumers of JISC service newsfeeds outside the UK academic community.
8. One or two small projects are commissioned to examine the flow of user information sideways between presentation services within the JISC IE and backwards between presentation services and content suppliers. The issues of stripping out personal information (possibly by a trusted intermediary - see recommendation 3.2 above) and possible restoring of that personal information (as results come back through the presentation service to the end user) should be examined.
9. Some research should be done into the uses of customisation and personalisation to extend access to disabled users. This could be done through a separate project or through integrating this concern into the brief of the other recommended projects above.