Skip to content

Commit b9c1c68

Browse files
committed
ADR 22: Preferred identifier algorithm
1 parent 2fb68b2 commit b9c1c68

File tree

1 file changed

+39
-0
lines changed

1 file changed

+39
-0
lines changed
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,39 @@
1+
# 23. Preferred Identifier Algorithm
2+
3+
Date: 2024-11-28
4+
5+
## Status
6+
7+
Draft
8+
9+
## Context
10+
11+
When moving to a model where multiple documents have multiple identifiers, we need to understand how we choose which one to display for human identification purposes.
12+
13+
## Decision
14+
15+
Scores for individual components are multiplied together; IDs scoring 0 are never routinely considered preferred. (Scores for a component are 1 unless otherwise specified)
16+
17+
#### Deprecation
18+
19+
Identifier is deprecated: 0
20+
Document is deprecated: 0
21+
22+
#### Schema
23+
24+
Default: 0.5?
25+
NCN: 1
26+
BAILII: 0.8
27+
FCL SQID: 0.01
28+
29+
#### Recentness
30+
31+
Take the most recent (which measure of recentness? on publication? identifier?) if all other choices are even.
32+
33+
### Other options
34+
35+
We probably want to be able to limit to "only public documents" (for the PUI), "only human friendly identifiers", which can also add a 0 modifier to an identifier/document pair.
36+
37+
## Consequences
38+
39+
We have a modular system for determining the best identifier/document to return in multiple scenarios.

0 commit comments

Comments
 (0)