Posts made by Christopher Teplovs

I found this article a bit dreary, but only because I dislike the focus on management (at the expense of focusing on learning). This is kind of a sore spot for me: I see a lot of financial resources going into learning management and not that much going into learning. Should we embrace this application of "learning" analytics, or should we actively push back against it in favour of focusing on a more traditional interpretation of "learning"?

Baker & Yacef say:

It has been proposed that educational data mining methods are often different from standard data mining methods, due to the need to explicitly account for (and the opportunities to exploit) the multi-level hierarchy and non-independence in educational data [Baker in press].

which I agree is confusing. So I searched for the forward reference (Baker, in press) and found the following:

Educational data mining methods often differ from methods from the broader data mining
literature, in explicitly exploiting the multiple levels of meaningful hierarchy in educational data.
Methods from the psychometrics literature are often integrated with methods from the machine
learning and data mining literatures to achieve this goal.

For example, in mining data about how students choose to use educational software, it may be
worthwhile to simultaneously consider data at the keystroke level, answer level, session level,
student level, classroom level, and school level. Issues of time, sequence, and context also play
important roles in the study of educational data.

(via a preprint of his chapter)

Ah, the joys of sorting out semantics in a multi-disciplinary field ;)

I work with a group that uses Knowledge Forum, which is generally held up as a "knowledge building environment". In the interest of fair disclosure: I'm currently the technical lead on the project. We maintain a lot of data about user behaviours, and we tend to write those data into the database itself. That sets us up for being able to use the recorded data as objects of inquiry themselves.

We have a long history of using analytics: our "ATK" (Analytic ToolKit) has been used for over a decade and provides a variety of reports. Please see for an overview.

We have also developed a series of "assessment applets" for Knowledge Forum that use those stored data and present them to the users. See for details about those.

The upshot of all this is that we find that feeding the results from the analytics back into the course really changes the dynamics. We've called this "concurrent, embedded and transformative assessment" in the literature about Knowledge Building.

Funny you should mention scrutable learner models: In Denmark I will be working with a large team that includes Susan Bull (a close colleague of Judy Kay's -- with luck we'll manage to get some of her attention too) and we will be focusing on Open Learner Modelling. My contribution to the project will be the implementation of a communication and negotiation layer to communicate the learner models to the interested parties (students, teachers, administrators, parents, etc.). The big challenge, as you rightly point out, is how to communicate the model of a complex thing (the learner model in this case) without using a complex representation. I hope we'll take up some issues around representations of complex entities in Week 4 with tools like Gephi.

I'd like to have an automatically-generated network graph or map that showed me various representations of all the posts in the moodle forums. I'd like to be able to arbitrarily reorganize the postings according to various criteria: authorship, chronology, theme, keywords, latent semantic content, relatedness, and so on. I'd like to be able to overlay additional information, like whether I had read the postings, atop the two- or three-dimensional display of the organizational variables. I'd like to be able to identify criteria that make certain notes or postings "rise up" and then be able to manipulate those postings.

You did say to dream ;)