Landscape

Personalisation and Socialization in the Lifelong Academy – the landscape for service development

You can navigate directly to the following sections or download the full document

A brief history

The past two decades have precipitated rapid development of opportunities to personalise the virtual experience, which might be characterized in three relatively distinct phases:

  • Interfaces – Emphasis in the 90s was predominantly on customisation of the interface and the desktop, building on metaphors popularized by Apple and Microsoft.
  • Services – Subsequently the emphasis shifted to services enabling the selection and filtering of services, such lists and feeds, according to preferences, productivity and platform.
  • Networks – The past 5 years has seen the emergence of new forms of large scale ‘socio-technical activity’, mediated through social and activity networks wherein personalisation is enhanced by following others and informed by web scale patterns

Next generation opportunities lurk in the promises of open data, open content and the Semantic Web.

Since 2006 JISC has commissioned a number of projects with a specific and sometimes exclusive brief to research and develop as personalisation in the context of the information environment, serving teaching, learning and research. Personalisation of learning has also been an important focus of the work of Becta (Harnessing Technology), Futurelab, and TLRP.

Meanwhile user awareness of potential for personalisation has been rapidly developed through widespread experiences of services such as social networking, shopping, entertainment and comparison sites. These experiences have traded personal disclosure against the benefits to be derived from activity, recommendation and location data. Popular tools and services have embedded a number of features in the group mind – for example, the desktop metaphor, browser-based access, favorites, the Facebook activity feed, awareness services (such as RSS Feeds), save for later & other lists.

However we must bear in mind that:

  • There is no single definition of personalisation (although we supply an all-encompassing definition later in this report)
  • Personalisation may have become a vacuous catchword through overuse in such as the 2008 Harnessing Technology Strategy
  • Generic applications and services are not necessarily transferable in to teaching, learning and research, which might require domain specific features

The current landscape

Here are some of the key debates and challenges regarding the scope of personalisation and its implications for systems and services in learning and research.

Dimensions – What does personalisation cover?

We can think about this in terms of:

  • What? The objects of personalisation – interface / metaphor, services, knowledge (metadata, content)
  • Why? The Motivations for personalisation – gathering, enhancement, reuse, presence, timing
  • How? The Means of achieving personalisation – machine automation / human selection, push / pull, passive / interactive
  • Who? The Ownership of the interactions and results – personal, group, public
  • When? The Persistence of personalisation – is it for life and therefore how portable is the content and data?

Paradigms – How can we conceptualize / model this space?

  • Personalisation meets Socialization (implying Architectures of Participation as it is not ‘single user’) – for example, am I creating my bookmarks alone or with the ‘help’ of a wider community?
  • Automation (Machines) v. Interactive (Humans) – for example, do I identify books or music of interest to me through searching or from recommendations (and are those recommendations generated by humans or by machines?)
  • Services (My corner of each service – e.g. my VLE or Twitter pages) v. Personal Space (My world incorporating multiple services – e.g. iGoogle)
  • Creator v. Consumer – In order to achieve the personal effects that I desire for my learning or research, do I wish to be active or passive, to collect or to share?

Audience – Do personal and social services evolve to suit audiences?

  • By definition, personalisation means widening blockquoteides / differentiating audiences (e.g. consider UG v. Research – Facebook v. Mendeley)
  • Researchers already have mechanisms and services
  • The primary opportunity is in targeting UGs

Affordable Futures – Is there an economic opportunity here?

  • The affordable future of learning is that learners take (more) control or responsibility in a (more) self-directed process
  • This may also be desirable in terms of pedagogy and employability skills
  • This future has to include Lifelong Learning, which implies portability of ‘my stuff’ and persistence of identifiers for content, courses and people
  • It requires us to rethink the role of platforms – VLE (our structure for your learning), Repository, PLE (Everything I need, Reflection), e-Portfolio (Evidence, Artefacts)
  • There are some responses bubbling up (e.g. see Sakai 3, Talis Aspire, Hedtek’s Manchester PLE) but there are major issues to address at sector level

Blockages – What stands in the way of delivering personalised services?

  • Engagement – people don’t personalise what they don’t like … and they don’t necessarily like what they can personalise
  • Content and data issues – silos, rights, portability
  • Visibility – especially in relation to the worlds users are already creating (e.g. in Facebook or with iGoogle)

Tools – Does the technology drive personalisation?

The interaction between user requirements pull and technology push is complex but there is no doubt that popular tools can be taken as a baseline proxy. Consider the following:

  • Top Apps for iPhone and iPad
  • Top tools for learning – see www.c4lpt.co.uk/recommended/ and www.readwriteweb.com/archives/top_10_web_apps_for_students.php
  • Desktop – iGoogle, Netvibes and other popular web start pages
  • Bookmarking – Delicious, Digg
  • Social Network Services – Facebook, Mendeley
  • Real-time Web – RSS, Twitter

A broader definition

That landscape suggests a broader definition of personalisation and its implications for the sector.

The following definition is sufficiently broad for our purposes here: ‘Personalisation is the supply of services and/or data based on a model of a user’.1

Within this definition, supply and model are interpreted broadly, to mean by a machine or human agent. Examples of human agency include personalised shopping, by one person for another, and self-regulated learning2, by someone for him/herself. Examples of machine agency include automated recommendations made by services such as Facebook (possible friends) and Amazon (possible purchases).

This demand / opportunity for personalisation is important to the JISC mission for a number of reasons, including

  • Users are familiar with such services and therefore have expectations
  • HE services within the Information Environment compete / interoperate with external services that offer elements of personalisation in areas such as resource discovery
  • Group activity is increasingly central to learning and research – and groups are simply a specialisation of the requirements placed on services by inblockquoteidual personalisation
  • The affordable future of learning may be dependent on users taking more control or responsibility for their learning and for the resources they use, share and, particularly, that they enhance in personal ways

Within that setting, strategic considerations arise about the content and data that is referenced or appropriated or generated by users – which we might call ‘my stuff’. If the sector is to take ‘my stuff’ and ‘our stuff’ (for groups) as serious parts of the scholarly and lifelong learning environment, there is a requirement to consider core processes with persistence in mind – such as Find, Reference / Revisit, Acquire / Copy, Repurpose, Group together, Share, Enhance.

Some Challenges

This section sets out the challenges faced in the development of personalisation (and social) services. It distinguishes between Challenges in hand A number of issues key to the design and delivery of personalised services are well provided for by services put in place by JISC:

  • Authentication and authorization – There are a variety of means to manage the information that underpins personalised services. Within the sector integration with Shibboleth or AthensOpen offer well understood capabilities.
  • Accessibility – TechDis promotes standards and offers strategies to ensure that personalisation can be engineered to overcome disability.
  • Legal – JISCLegal is able to offer guidance on requirements regarding consent and obligations under the Data Protection Act. This is increasingly important as personalisation takes place in a social context; for example, where the user’s data contributes to rankings and recommendations.

In a rapidly development global knowledge economy, these issues will always represent areas of risk and complexity. They are not unique to the HE environment – they face anyone developing such services. The sector is very well served, managing risk and addressing challenges above and beyond what might be expected of its commercial providers and its competitors. Outstanding Challenges There are other issues of responsibility and service assurance that are less well understood and less comprehensively addressed, notably:

  • Security – Is personalised content secure in terms of backup (i.e. taken as seriously as metadata or corporate content)? Given the user’s permission, is it secure for the future benefit of the UK scholarly community (i.e. not lost in the global data space)? To what extent this matters is variable, bearing in mind it may range from trivial configuration parameters to a significant volume of intellectual content (e.g. annotations, bookmarks).
  • Longevity – Can providers give reasonable assurances about the longevity of their service offer?
  • Availability – Is the service level fit for purpose, typically 24*7 on any device in the case of personal services?
  • Control and access – What controls are in place regarding how personalised content is used / reused in services in the UK and elsewhere in the world?
  • Veracity – Is there any method of assuring the veracity / integrity of personal contributions upon which some services depend for such as recommendations?
  • Data portability – Can users take their personal data with them when they move between services and / or between institutions? (e.g. Export on progression and import in to a new system). This is potentially the most serious issue.
  • Ownership – is ownership vested appropriately, e.g. is a user’s attention data their own to do with as the user deems fit?

Whilst there are likely to be risks that users are willing to accept as part of living and working online and about which they are unlikely to have taken conscious decisions, it is important to assess them from the sector perspective – as the service provider in the specific context of Teaching, Learning and Research. Integrate, Interface or Ignore? Our response to these service challenges may help to answer the underlying questions of whether HE systems should

  • Integrate personalisation locally or above campus – potentially though not necessarily using third party tools (See Sakai 3 plans for example)
  • Interface with external services (outside sector and often global services)
  • Ignore the issue and let users choose their own services (e.g. use Delicious) in the same way they select desktop tools for tasks such as word processing

Major service examples to consider in addressing these options include bookmarking services (such as Delicious and Digg) and Social Networks (such as Mendeley and Facebook).

There is no general answer. Should, for example, inblockquoteidual institutions or the sector at large should take on the development of an ‘architecture of personalisation and / or participation’? The market is too immature to generalize such decisions – unlike, for example the enterprise database (DBMS) market where no HEI should have built their own post 1985. The availability of personal and social services may reach that position in the current decade, but right now the sector and its institutions must respond on a case-by-case basis.

Here are some examples (the product names being incidental):

  • Should undergraduate collaboration take place in Facebook, in a locally accessible instance of Ning or inside a VLE or institutional PLE?
  • Should students manage their personal records of experience / portfolios in a blog, in the licensed copy of Pebblepad or on their desktop?
  • Should a distributed research group develop its network through Mendeley, the institutional VRE or using the homegrown tools linked to their experimental data?
  • Should students develop their references and annotations in a local RefWorks instance (tightly integrated with both LMS and VLE) or using Zotero in the cloud and engaging with its wider community?

The key questions that should inform case-by-case decisions to build or to buy, to localise or to trust to the network, are:

  • Is it safe? – as per the considerations above
  • Is it fit for purpose? – bearing in mind the specifics of TLR, including assessment and IP protection
  • Is it convenient? – reliable, easy to access, open for personal use, the user’s data to keep
  • Are there beneficial ‘network effects’? – scale of community enhances personal feedback loops
  • Is it affordable? – can the institution or the sector provide the required / expected functionality?

1 Mark van Harmelen, see http://hedtek.com/?p=342

2 Zimmerman, BJ, Bonner S, and Kovach, R. Developing Self-Regulated Learners Beyond Achievement to Self-Efficacy. Psychology in the Classroom: A series on Applied Educatiuonal Psychology. American Psychological Association. APA Books. 1996.