Initial readme
Browse files
README.md
CHANGED
|
@@ -1,3 +1,15 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc0-1.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc0-1.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
pretty_name: JFK declassified documents March 2025 cleaned
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
On March 18th 2025, 2182 previously classified documents related to the investigation the assassination of President John F. Kennedy were made public at https://www.archives.gov/research/jfk/release-2025
|
| 9 |
+
|
| 10 |
+
This dataset contains a "cleaned" version of the `zlatinb/jfk-2025-raw` dataset. I've used `cleanup.py` to perform a best-effort cleanup based on some heuristics.
|
| 11 |
+
|
| 12 |
+
The columns in the dataset are:
|
| 13 |
+
* `File` - name of the original document that you can download from the government website
|
| 14 |
+
* `Accepted` - the lines from that document, prefixed with line numbers, that passed a certain threshold of "cleanliness"
|
| 15 |
+
* `Rejected` - the rest of the lines of the source document
|