Menu

[self-promotion] All 468,352 Library of Congress Subject Headings (LCSH) as MARC records, rebuilt weekly

#datasets#redditdatasets#marc#every#week
Reading 0:00
15s threshold

The LCSH are the vocabulary libraries around the world use to describe what books are about. LC has maintained them since 1898, and gives them away in bulk, but only as linked data (JSON-LD and N-Triples). MARC is the format library systems actually import, and LC sells it in bulk. There was no free way to get the whole set as MARC, so I made one.

The mirror is here: https://bibliotecaalonsocossio.com/lcshmirror/

What’s there:

The full set every week: 58 MB gzipped MARC 21, 56 MB gzipped MARCXML (~570 MB uncompressed), plus a CSV of every record (LCCN, last-change date, heading).

Change files per release: just the records new or changed since last week, plus a list of deleted headings.…

Continue reading — create a free account

Join HashtagPLUS to read full articles, follow hashtags, vote, and join the conversation.