Adaptive technologies for learning

From Cognitive Accessibility Task Force

Learners are individuals, and as such, each of them is different. Computer- and web-based educational research is slowly starting to take that into consideration, moving away from the “one-size-fits-all” approach of educational broadcasting. The move is faster in the research field, where many adaptive educational systems already take into account different learner features such as goals/tasks, knowledge, background, hyperspace experience, preferences and interests.

The educational technology research in the area of adaptivity dates as far back as the 1970s, starting from the SCHOLAR system for South American geography education (Carbonell, 1970). The general thrust of the earlier adaptivity movement was in the form of intelligent tutoring systems (ITSs) to apply artificial intelligence to mimic human tutoring with the hope of achieving significant learning improvement similar to the two sigma improvement documented with one-on-one human tutoring (Bloom, 1984). While ITSs did have some success in the 1980s and 1990s, they proved to be too expensive for many educational applications and not well suited for supporting ill-structured problem solving domains. Meanwhile, new technologies have emerged and new ways to leverage the lessons learned from earlier adaptivity applications have come into existence in the context of personalized learning. Personalizing education is the first grand educational challenge mentioned in A Roadmap for Education Technology (Woolf, 2010). Technologies now available to support personalizing learning and instruction include dynamic user modeling, real-time data-mining, adaptive and intelligent systems, gaming environments, mobile and ubiquitous technologies, and more. The 2011 New Media Consortium’s Horizon Report (http://www.nmc.org) mentions nearly all of the same enabling technologies.


Need for adaptivity

  • Learners generally learn on their own without external support.
  • Same learning systems are used by variety of learners from all over the world, who possess different cognitive and behavioural characteristics.
  • Customised system behaviour helps in reducing meta-learning overhead for the learners and allows focus on the completion of actual task.


Spectrum of adaptive learning systems

  • Systems that allow the user to change certain system parameters and adapt the system behaviour accordingly (fully adaptable systems)
  • Systems that facilitate user-desired adaptability which is supported by tools and performed by the system
  • Systems that enable users to select adaptation from system suggested features
  • System that initiate adaptivity with pre-information to the user about the changes
  • Systems that adapt to the users automatically based on system’s assumptions about user needs (fully adaptive systems)


How does adaptivity work?

  • System monitors user’s action patterns with various components of system’s interface.
  • Some systems support the user in the learning phase by introducing them to system operation.
  • Some systems draw user’s attention to unfamiliar tools.
  • User errors are primary candidate for automatic adaptation.


Inputs for adaptivity in learning

Adaptivity through the dynamic learner modeling

  • Performance based model
  • Cognitive trait model
  • Learning styles


Adaptivity through the use of location modeling

  • Location based optimal grouping
  • Location based adaptation of learning content

(Technologies commonly used to identify location: GPS navigation data, Cellular Network Base Station data)


Adaptivity as per surrounding environment
(Technologies commonly used to identify surrounding environment: Public databases of Points of Interests (POIs), QR Codes, Wi-Fi & Bluetooth Access Point, Active and Passive RFIDs)


Adaptivity through the identification of technological functionality

  • Identifying various device functionality
  • Dynamically optimize the content to suit the functionality

(Technology functionalities commonly used: Display capability, Audio and video capability, Multi-language capability, Memory, Bandwidth, Operating platform)


Adaptivity through the use of surrounding context

  • Identifying specific context-aware knowledge structure among different domains
  • Identify the learning objective(s) that the learner is really interested in
  • Propose learning activities to the learner
  • Lead the learner around the learning environment


Cognitive Trait Model in detail

Cognitive trait model attempts to create profiles of learners in terms of their cognitive traits such as working memory capacity, inductive reasoning ability, etc. in order to allow learning systems to provide adaptivity accordingly (Lin, 2007).

  • Working Memory Capacity: allows us to keep active a limited amount of info (7+/-2 items) for short time (Miller, 1956).
  • Inductive Reasoning Ability: is the ability to construct concepts from examples.
  • Information Processing Speed: determines how fast the learners acquire the information correctly.
  • Associative Learning Skill: is the skill to link new knowledge to existing knowledge.


Examples of adaptation based on cognitive traits: Case of low working memory capacity

  • Number of paths should decrease in order to protect learners from getting lost in too much information, from overloading the working memory with complex hyperspace structure.
  • Relevance of paths should increase in order to give learners important information directly without irrelevant information.
  • Amount of information should decrease so that only important information is given to the learners in order to protect them from information overload and to give them more time to review essential content if necessary.
  • Concreteness of information should increase so that the learners can grasp the fundamental concepts first and use them to generate higher-order concepts.
  • Number of information resources for same subject content should increase, so that the learners can choose the media resources that work best along their cognitive styles.


Characteristics of Cognitive Trait Model

  • be transferable across different domains;
  • be persistent and stays valid for long durations;
  • be used to provide adaptivity to suit individual learner’s cognitive capacity;
  • be used to supplement existing performance-based models;
  • automatically select appropriate cognitive theories for different individual; and thus,
  • be a life-long learning companion.


References

  • Bloom, B. (1984). The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring. Educational Researcher, 13 (6), 4-16.
  • Carbonell J. R. (1970). AI in CAI: an artificial intelligence approach to computer-assisted instruction. IEEE Transactions on Man-Machine Systems, 11, 190-202.
  • Lin, T. (2007). Cognitive Trait Model for Adaptive Learning Environments, PhD Thesis, Palmerston North, New Zealand: Massey University. Retrieved 4 June 2014 from http://mro.massey.ac.nz/handle/10179/1451.
  • Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63 (2), 81–97.
  • Woolf, B.P. (2010). A roadmap for education technology. Retrieved 4 June 2014 from http://www.cra.org/ccc/files/docs/groe/GROE%20Roadmap%20for%20Education%20Technology%20Final%20Report.pdf.


Further readings

  • Brusilovsky, P. (1996). Methods and techniques of adaptive hypermedia. User Modeling and User-Adapted Interaction, 6 (2-3), 87-129.
  • D’Mello, S., Craig, S., Fike, K., & Graesser, A. (2009). Responding to learners’ cognitive-affective states with supportive and shakeup dialogues. International Conference on Human-Computer Interaction (pp. 595 - 604 ). Springer, Lecture Notes In Computer Science; Vol. 5612.
  • Graf, S., & Kinshuk (2014). Adaptive Technologies. In J. M. Spector, D. Merrill, J. Elen, & M. J. Bishop (Eds.), Handbook of Research on Educational Communications and Technology, Heidelberg: Springer, 771-779.
  • Kinshuk, & Lin, T. (2003). User exploration based adaptation in adaptive learning systems. International Journal of Information Systems in Education, 1 (1), 22-31.
  • Kinshuk, & Lin, T. (2004). Cognitive profiling towards formal adaptive technologies in web-based learning communities. International Journal of WWW-based Communities, 1 (1), 103-108.
  • Popescu, E. (2010). Adaptation provisioning with respect to learning styles in a web-based educational system: An experimental study. Journal of Computer Assisted Learning, 26 (4), 243-257.