ContextInformed Adaptive Hypermedia

1
Context-Informed Adaptive Hypermedia

• Alexander OConnor
• Owen Conlan
• Vincent Wade
• oconnoat, Owen.Conlan, Vincent.Wade_at_cs.tcd.ie
• Knowledge Data Engineering Group
• Trinity College, Dublin

2
Adaptive Hypermedia for eLearning

• Developed from Intelligent Tutoring Systems (ITS)
  and Hypertext
• Adaptive Hypermedia1 systems compose content
  based on rules and course design with reference
  to model of learner
• Models are generally highly detailed.

3
APeLS2

• Adaptive Personalised eLearning Service
• Multi-Model Metadata Driven AH System
• Uses Jess to build an XML document of the course
• Narrative compares attributes of the Content and
  the Learner Model
• Content is referred to indirectly
• Candidate Groups

4
APeLS Architecture
Narrative Models
Content
Learner Models
5
Context for Adaptive Hypermedia

• Context in Adaptive Hypermedia is composed of a
  variable set of axes with the following
  properties
• Not core model components
• Potentially interesting to the system
• Context has deep and shallow models
• Context for one system might not be context for
  another

6
Why Add Context?

• AH systems already have methods for modeling
  relevant data.
• These methods are tailored and effective
• But, the models tend to be complex
• Need some way to handle extra concerns easily
• Define Context as data that would be useful, but
  is not core to the AH

7
Context-Informed AH

• Context supports additional concerns for the AH
• Factors not specified when the course was created
• A Context Interpreter translates these extra
  factors into a known vocabulary
• This is done by providing mechanisms to pass
  information about the state of the narrative and
  models to the CI, which can make changes
• Decisions on a list of concepts/entities passed

8
Context-Informed APeLS
Narrative Models
Content
Learner Models
9
Mechanisms

• Complete Model Enrichment
• Pass the contents of a model to the Adaptive
  engine, which alters it
• User Model Update
• Selected Model Enrichment
• Context decides on the membership or order of a
  portion of the model from a list provided
• Candidate Group Manipulation
• Collaborative Dialogue
• Define Decision points which are answered by
  Context
• Narrative Choice

Different Mechanisms impose different
requirements for shared knowledge
10
Advantages

• This method permits Adaptive Hypermedia to make
  use of a wider knowledge set.
• Without having to model it directly
• Increases Adaptivity
• Provides Interoperability framework
• The integrity of AH core deep models is
  maintained
• While the CI is able to employ shallower
  techniques

11
Conclusions

• Separated Architecture
• Core concerns are modeled deeply by Adaptive
  Hypermedia
• Context handles extra inputs separately
• Translated to terms known to the Adaptive Engine
• Applications to other systems

12
References

• Brusilovsky, P. Methods and techniques of
  adaptive hypermedia. In P. Brusilovsky and J.
  Vassileva (eds.), Spec. Iss. On Adaptive
  Hypertext and Hypermedia, User Modeling and
  User-Adapted Interaction 6 (2-3), 87-129
• Conlan, O. Wade, V. Bruen, C. Gargan, M.
  Multi-Model, Metadata Driven Approach to Adaptive
  Hypermedia Services for Personalized eLearning.
  Second International Conference on Adaptive
  Hypermedia and Adaptive Web-Based Systems,
  Malaga, Spain, May 2002