Podcast Research

The Podcast Relational Database Project

A little over a year ago (~ June 2005) I discovered podcasts. I started out slow, but I now consistently listen to 18 different podcasts. Even though I listen to a lot of shows, I need more to fill my podcast appetite. My problem is finding new shows. I tried using podcastalley, but I lost faith in them after I listened to one of the top two shows (it was either pottercast or mugglecast) and was appalled by how horrible the show was. What I needed was to find people who listen to similar shows and discover what shows they like that I am not currently subscribed to. I was convinced that someone had already gathered this information and had a nice site put together for it, but alas I was unable to find the information I was looking for. So, I decided to put the site together myself.

The idea behind this project is to first find out what groups of shows people listen to and build a relational database to contain this information. Once the database has enough information, a user can provide a list of shows they listen to and get back a prioritized list of additional shows they would likely be interested in. It is simple in concept, but I know that it will involve quite a bit of my time to develop the algorithm to construct the suggestion list. Because I didn't want to waste my time on a project that no one is interested in I have broken it into phases with the idea being that interest in each phase will determine how the project moves forward. I currently have three phases planned. If you have any questions about this site and any future plans, feel free to contact me.

Phases

Phase 1
The first phase's purpose is to build as large of an initial database as possible. A survey that is open to the public gathers a list of podcasts a person listens to. This phase is planned to last for two months, or until 500 surveys have been taken. When Phase 1 ends, the public survey will be closed.

Phase 2
Assuming Phase 1 is successful, Phase 2 will be the development of an algorithm to suggest podcasts of interest to users based on their current list of shows. In order to minimize spamming and data corruption, users will most likely be required to login securely to view this information and to manage their podcast lists. The deliverable of Phase 2 is the prioritized podcast suggestion list.

Phase 3
This final planned phase is the "bell and whistle" phase. Phase 3 will consist of the addition of features to improve the functionality of the Podcast Research site. Current thoughts for this include direct links to rss feeds for podcasts, reviews for podcasts, and maybe even a forum. Most likely, these developments will be user driven with high demand features given priority.