Sloan-C Conference

I’ve had the pleasure of presenting at the Sloan-C conference this past Friday in Orlando. I thought I would share the basic premise of my presentation here on my blog.

PocketKnowledge Learning Principles: Overview

The PocketKnowledge interface was designed primarily to: a) promote meaningful learning by facilitating the interconnection of conceptual knowledge spaces, and b) illustrate the contours of knowledge production within the field, as it is produced locally at the institution. The following discussion will focus on the first element, or how learning was promoted through the interconnection of conceptual knowledge spaces.

Connecting Knowledge Spaces

One of the first considerations in designing the PocketKnowledge interface was to understand the ways in which people use the web. A useful study for illustrating how people use the web was conducted by Sellen, Murphy and Shaw (2002) who looked at the how knowledge workers use the web. In this study, they found that use was divided into six main categories, including:

  1. Finding: using the web to find something specific (ie- finding a phone number, how to spell a word, retrieving a specific document)
  2. Information gathering: Goal oriented search for information; however, less specific than “Finding� (ie- looking for a job, researching different products before making a purchase)
  3. Browsing: Web browsing with no specific goal in mind (ie- entertainment, staying informed)
  4. Transacting: Use the web to make a transaction (ie- bank transaction)
  5. Communicating: Using the web in chat rooms or discussion groups
  6. Housekeeping: Using the web to check the accuracy and functionality of web resources (ie- check that links are working)

The study found the following distribution of web activity: 24% Finding, 35% Information gathering, 27% Browsing, 5% Transacting, 4% Communicating, and 5% Housekeeping (see figure 1). Although knowledge workers may not typify an academic community in all respects (i.e., academic communities have a large demographic divide between students and faculty), the study is useful in that it divides web-browsing as a general category into more specific terms which more accurately reflect human activity. For example, the activity associated with retrieving a specific document on the web is a much different activity than non-goal oriented browsing. The study is also useful in that it highlights a range of web use which is not specifically goal oriented. In the case of an institutional digital repository, these could include the types of features not necessarily associated with a search and retrieval paradigm, such as connecting users with related materials he or she may not have specifically requested. In effect, uses for a system which expand beyond search and retrieval, such as functions for staying informed, entertained, or taught something new, are suitable aims for any web-based system, especially considering that much of web-use (~30%) is of this nature. Further, we speculate that this range of use, which would tend to be more unencumbered, is a prime opportunity for learning.

One method to make less goal-oriented times into moments for learning is to subject the web user into encounters with materials he or she never explicitly requested. Further, the web user is more likely to comprehend new content if it is semantically related to content he or she is already familiar with. According to Ausubel’s assimilation learning theory (1968), meaningful learning is a process in which new information is related to an existing and relevant aspect of an individual’s knowledge structure. The notion of integrating new knowledge with pre-existing knowledge forms much of the basis for educational knowledge maps and concept maps (O’Donnell, 1998; Novak, 2002). In terms of system design, this would mean that a system should first present the user with something he or she knows, and then provide the user with additional information which is highly connected to their existing knowledge structures. In PocketKnowledge, the browsing interface was designed to do just that. The interface was designed not to look like a knowledge map, bur rather to induce a knowledge map in the mind of its user through the user’s interaction with a highly linked network of materials. The browsing interface connects knowledge products with related knowledge productions via tag, pocket, author and uploader. The user can jump amongst and discover connections between these categories. For example, figure 2 illustrates a network representation of a user’s traversal of materials. Although PocketKnowledge’s browsing interface is not explicitly a knowledge map, upon inspection of a user’s page traversals, one can see that a knowledge network is formed. In sum, we posit that when a learner navigates an interconnected, networked space, he forms a mental model of the conceptual space. This mental model can be mapped onto pre-existing knowledge structures. This process is more efficient if the learner has knowledge of one or more parts of the knowledge map which he could use to relate new knowledge to.

Also, we posit that navigating a network (or experiencing it incrementally), and slowly integrating its structure into preexisting knowledge structures, is more cognitively effective than attempting to understand a completed network diagram. We are currently testing this hypothesis in a series of usage experiments. The basis for such a hypothesis resides on two notions. The first of which is the observations that humans have difficulty processing visualizations which include more items than can be stored in short-term memory. For example, Kosslyn (1989) finds that graphs should include a relatively small number of elements in order to be effectively processed by the human mind (pp. 191). The second notion guiding this design decision is the cognitive predisposition to create mental models to describe the interrelationships between various phenomenon (Johnson-Laird, 1983). Hence, it is not necessary to show all relationships at once, in the case of a network visualization, but rather to expose the user to all the relationships eventually and at the user’s own pace.


Ausubel, D. P. (1968) Educational Psychology: A Cognitive View. New York: Holt.

Kosslyn (1989). Understanding Charts and Graphs. Applied Cognitive Psychology, 3, 185-226.

Novak, J. D. (1998) Learning, Creating, and Using Knowledge: Concept Maps as Facilitative Tools in Schools and Corporations. Mahway, NJ: Lawrence Erlbaum.

O’Donnell, A.M., Dansereau, D.F., & Hall, R.H. (2002). Knowledge Maps as Scaffolds for Cognitive Processing. Educational Psychology Review, 14, 71-86.

Sellen, A.J., Murphy, R. & Shaw, K. L. (2002). How Knowledge Workers Use the Web. Conference on Human Factors in Computing Systems: Proceedings of the SIGCHI conference on Human factors in computing systems: Changing our world, changing ourselves, April 20-25, (pp. 227-234), New York: ACM Press.

Leave a Reply

Your email address will not be published. Required fields are marked *