2011 WMMC Problems
Problem 1: Not Your Father’s Textbook
College textbooks have drastically changed over the past decade. In particular, new textbook formats and a variety of purchasing options have forced students, publishers, authors and book stores to adapt. What will the college textbook market look like ten years from now?Problem 2: In Yo’ Facebook
Online social media companies such as Facebook, Twitter, and Digg don’t manufacture goods in the traditional sense, yet all these companies
produce an extremely valuable product; preference focused user-data. The ability of web-based companies to understand their data is what
makes it a commodity of interest; specifically because it allows potential advertisers the ability to create marketing strategies targeted
to particular subsets of the population.
You have been provided with a two week snapshot of data from a page rank website that allows users to identify general links of interest
(organized into a variety of categories and subcategories). Users of this site have the ability to comment on the listed articles and respond
to fellow users comments. Moreover, all items on the site (stories, comments and replies) can be liked or disliked by any user. You have been
hired by the social media company that owns this data to identify a general strategy for determining users who wield the most influence on
the site as a whole and/or within different categories and subcategories that might be of interest to advertisers. The company is hiring you
because they have learned through experience that the most “liked” users have not necessarily been the most influential. As a result they are
looking for a general approach to solve the “influential user problem”. The company is also interested in updating their data collection
capabilities; what additional data, if any, would make the influence of users more easily identifiable.
The complete data set for this problem is available here.