Skip to content

Latest commit

 

History

History
94 lines (66 loc) · 3.52 KB

README.md

File metadata and controls

94 lines (66 loc) · 3.52 KB
Raw

All Podcasts Dataset

This is a free dataset of (almost) all publicly available podcasts - at least the ones that I could find that were actually working and at least relatively well formatted.

This dataset consists of ~135,000 podcasts. Each entry was generated by getting the RSS or Atom feed for a podcast, crawling it and then capturing whatever information was available. The data was captured in August 2014.

The data is in tab-separated (.tsv) files that should be easy to import into pretty much any system. Each file contains all the podcasts that start with that letter. In the .tsv files, fields containing quotes in the data are quote delimited. Empty fields are represented as "". Pretty simple stuff.

Why?

I was doing an experimental project where I needed a database of all podcasts. Since someone else may be looking for the same thing, I thought I'd share.

Things you could do:

  • Build a podcast directory for your podcast player app
  • Do some machine learning with the data
  • Revel in how many RSS feed urls you now know
  • (Please don't) spam all the email addresses of every podcast owner

Restrictions and Limitations

All this data comes from RSS / Atom feeds published by the podcast authors, so I don't own any of that content. Do whatever you want. But if you do something cool, sent me a tweet at @ageitgey or drop me an email and show me what you built.

Since this data was built for a quick hack, I make no warranty that it is complete, accurate or anything else. There's probably some amount of bad data. But it worked well for my needs.

Data elements

Each row in the .tsv represents one podcast.

Each row contains the following fields (in this order):

slug

A computer-generated short name or "permalink" for the feed (you can ignore this).

name

The name of the podcast, as reported in the RSS feed.

image_url

A url to a cover image for the podcast.

feed_url

The url of the RSS / Atom feed itself that was crawled.

website_url

The homepage of the podcast, as reported in the RSS feed.

itunes_owner_name

The podcast owner's name, as reported via itunes:owner tags in the RSS feed.

itunes_owner_email

The podcast owner's email address, as reported via itunes:owner tags in the RSS feed.

managing_editor_name

The managing owner of the podcast, as reported in the RSS feed (often missing).

managing_editor_email

The email address of the managing owner of the podcast, as reported in the RSS feed.

explicit

Whether or not the feed claims to contain explicit content as reported in an itunes:explicit tag.

description

The description of the podcast, as reported in the RSS feed.

itunes_summary

The iTunes-specific description of the podcast, as reported in a itunes:summary tag in the feed.

Example entry

Here's one podcast as it appears in the data set:

slug name image_url feed_url website_url itunes_owner_name itunes_owner_email managing_editor_name managing_editor_email explicit description itunes_summary
my-brother-my-brother-and-me My Brother, My Brother And Me http://assets.libsyn.com/content/7416218.jpg http://mbmbam.libsyn.com/rss http://www.mbmbam.com Justin McElroy [email protected] [email protected] [email protected] true Free advice, from three of the world's most qualified experts. My Brother, My Brother and Me is an advice show for the modern age.