BIBFRAME The Future of Linked Data in Libraries: Assessing BIBFRAME Against Best Practices

Posting to Bibframe

On 1/23/2016 2:24 AM, Martynas Jusevičius wrote:

I encourage participants of this list to watch the following > presentation by Robert Sanderson before diving into philosophical > discussions about events in spacetime:
https://www.youtube.com/watch?v=2-U-Qd37WgE

Thanks for that. Very interesting.

The discussion about reuse is at: https://youtu.be/2-U-Qd37WgE?t=31m26s, and I agree that if you want developers to include your information (which is the idea of linked data), it is necessary to reuse code whenever possible. Otherwise, you are creating a huge additional hurdle for the developers and many will choose not to include your information in any tools they create.

Of course, the problem with using other codes is that you end up stuck in certain ways if they define things differently than you do. So, that is why in my previous message I gave the definitions of “event” as used in other implementations. None of them use anything that is similar to any library usage. Therefore, if we are to make something that will be useful and valid in an “events” semantic system created by somebody else, we have to decide how to index our information to match others’ ideas of events.

If an event is seen only as a meeting of people, then in addition to the 111s, there will also be many, but far from all, 110s (e.g. 110 2_ |a Air Traffic Control Association. |b Annual Meeting |n (22nd : |d 1977 : |c Las Vegas, Nev.)) and even many 151/110 jurisdictional names (e.g. 110 10 |a United States. |b Congress |n (101st : |d 1989-1990))

If an event is supposed to cover other “things that happen” caused by humans such as wars and battles or events caused by non-humans, the new index will need to include some, but not all, 150s (150 __ |a Hurricane Katrina, 2005) and 151s (151 __ |a United States |x History |y Civil War, 1861-1865).

With subjects, there is the additional complexity that there are subdivisions, e.g. the subject “United States–History–Civil War, 1861-1865–Bibliography”. Would this still be considered an event?

If not, there is the URI for this subject, which includes the entire string. http://id.loc.gov/authorities/subjects/sh2007100437.html. Would this entire URI still be considered an event? If not, how would it work vis-a-vis the URI for “United States–History–Civil War, 1861-1865”?

Many others view events as individual performances as for instance, when someone wants to buy or sell tickets to a musical concert. Libraries don’t code to this level and you need to look at the entire record for individual performance information.

No matter what, it sounds like adding “event” would mean a major change for our current data.

-82

Share