Course Customization and Learning Social Networks

MOTIVATION: Easy access to freely available, educational content on the Web is challenging universities to use and adapt the World’s educational resources to provide a better onsite education, and just as importantly, challenging all of us to contribute to the World’s high-quality, freely available educational content. There are many dimensions to these challenges. One, for example, is quality control. There has historically been some debate over the quality of Wikipedia articles, though many of us use Wikipedia regularly, even if the vast majority of us don’t contribute to it. What can we do to contribute content and to vet content, on Wikipedia and other media? Other challenges have to do with insisting on physical presence and rising student expectations that education should be embedded in media that is accessible anytime and anywhere.

Project 2 asks you to think deeply about some aspects of your educational experiences and trajectories, and to reason about the educational resources around you so that you might design your own curriculum and suggest curricula for others.

PROJECT OVERVIEW: In particular, you are to fully specify, design and implement a database that supports an education-oriented social network. The system — database and primitive front-end — will be part student record system; part curriculum development tool; part recommender system; parts that are left to you to imagine, specify and implement; and all embedded within a social network that allows for various kinds of sharing amongst confidants (or any other synonym that you want for ‘friend’), while also protecting personal information as stipulated by individual users.

You might start this project by watching a panel discussion on the “unbundled university” (, as well as Doug’s Chronicle of Higher Education post at , and ITHAKA conference panel talk involving customized courses at, starting at about 26:40 into the video; slides at ITHAKA Presentation-10-16-12 ) and discussing as a group. Consider how you use educational materials, to include online material, but also conventional course materials, like textbooks and assigned readings. How do you meld these into a course of study? Presumably, your strategies will vary and presumably your database design would support a large space of individual preferences.

USE CASE SCENARIOS: Here is an example of the kind of information that your system might support (as realized by one example “super” student).

Jungsoon is taking a Database class at Some University, with a particular textbook (which is available in hardcopy or can be rented online), from a particular instructor. Jungsoon doesn’t like the textbook, but generally reads some of each chapter before turning to online videos on the Web and relying on instructor powerpoint slides that are posted on the Web. In addition to a favorite video series (, Jungsoon has found a couple of Wikipedia articles (e.g., on SQL querying and security) to be particularly helpful, as well as a few Khan Academy ( snippets as tutorials on summations that helped with understanding a particular theoretical aspect of query optimization. Her instructor has also posted a couple of isolated videos on B+ tree and extendible hashing indexing, posted to the Web. She has also found other resources to be less than ideal, ranging from less than optimal to worthless to confusing. She was so discouraged by the confusing treatment of a certain UML construct that did her own 10 minute video tutorial on it, and of course she has found an aggravating typo in her textbook, confirmed later by her instructor, but which caused no end of confusion when she first read it.

What would be required of a database to support Jungsoon in recording her network of material for the Database class, the sequencing of it, and her opinions about material on this network. How would a database intended to support this, represent certain RESOURCES like textbooks, courses, instructors, online material of various sorts? How would it represent meta data like Jungsoon’s opinions and other recommendations (e.g., how Jungsoon feels that the various material, central to her Database course and otherwise, should be networked together)?

Rather than Jungsoon’s personal database for her Database class, however, the database and system that you specify, design, and build, should support larger networks as well — networks of courses and networks of USERS. In particular, if Jungsoon recognizes a conceptual link between a topic in the Database course and a topic in an earlier AI course, some kind of explicit link between these might be represented. Jungsoon also has confidants at Some University, as well as other universities and “independents”. They are taking other courses, many on the same topic, and these confidants want to represent linkages between courses (and the network of material that they have build around these courses). Some users have even designed their own “courses” in different topic areas, by piecing together online material, to include online lectures, papers, texts, tutorials, and the like.

The social network might also allow users to enter grades received in classes, ratings of courses and instructors, and the like. One function that we might choose to implement, or at least discuss, is an ability of a user to access course ratings, and average grade received (from those who have posted grades), and course rating as a function of grade received (e.g., through a linear regression) — we will probably talk about such a function as a way of motivating a discussion on privacy preserving statistical analyses of databases generally. It might also be desirable for users (e.g., “students”) to point at (e.g., augment) the content of other users (e.g., “instructors”), and in some cases, not allowing a member of the latter category to see the pointers and/or not identify the user who is doing the pointing – thus, consider whether “confidant” relationships can be asymmetric (uni-directional)

POSSIBLE DATABASE ATTRIBUTES: Here are some items that your database might reasonably capture — there will be information about USERS and RESOURCES and connections (both through associations and abstractions, such as “TOPICS”) between them.

Names, addresses, aliases, etc

Universities attended, other schools attended, student status, job status

Courses taken, whens, wheres, instructors

Course names, descriptions, content, instructors, semesters/quarters/years offered, course and instructor ratings and recommendations, textbooks, other materials – “courses”, broadly construed, can be represented as a network of content sequencing of content

Courses prerequisite relationships within a curriculum, Course Equivalences between Universities (e.g., those that would transfer for each other), other relationships between courses, within a University and across Universities

Online material relevant to courses — type of material, location of material, on web or otherwise

Users have confidants, perhaps one way – followers that can’t be seen; there may be different kinds of users, or more likely different kinds of relationships between users

Others ….

SCHEDULE: Each remaining week, each group will discuss their ongoing design, in class and for the class. Each group will be given a 10 minute slice of time per class, with the intention that cross-fertilization of ideas occur across groups. Each week’s deliverables will be promulgated ahead of time, and the order in which groups will spotlight their designs will be promulgated ahead of time, so that no member of the group presenting first will be late.

By each Monday night, you will also upload an individual summary of your activities on the design for the prior week, including a general log of hours that you spent on the project and the design activities that you spent them on. You may also offer your assessment of your contribution to the group effort for that week, together with an opinion of how well the group is working together.