Amy L. Grude

What we did

Once we had chosen iTunes as our project topic, we created 3 personas, each representing a different type of user. We then used these personas to conduct a cognitive walkthrough and heuristic analysis of many typical tasks. From this data, we identified two areas in which iTunes does not support the way the users actually work and play with music.

One area we identified was contextual listening and interruptions. For example, a user might be listening to their iTunes library while writing a paper. Though they like all of the songs in their library, only certain songs are appropriate for homework. If a particularly loud and fast song comes on which distracts the worker, they will stop what they are doing, pull iTunes to the foreground, change the song, and then return to what they had been doing previously. We designed a system that would "learn" what the user prefers to listen to in different contexts (using open programs on the computer, time of day, ambient noise, etc) and allow them to teach those preferences to the computer using keystrokes that do not take them away from their current work. The other system we developed uses a recommender system and back-end semantic database to allow for a rich browsing environment.

In this project, I lead the group in ironing out the details of the systems, making sure we addressed how specifically the users would accomplish tasks. Additionally, I encouraged the group to think about how the user would move both forward and backward in the system and how they would view content in different modes. I also created the browsing visualization mockups.