Upload folder using huggingface_hub
Browse files- README.md +24 -14
- articles.parquet +2 -2
- nodes.parquet +3 -0
README.md
CHANGED
|
@@ -30,21 +30,31 @@ over-extraction.
|
|
| 30 |
|
| 31 |
| File | Rows | Description |
|
| 32 |
|---|---|---|
|
| 33 |
-
| `articles.parquet` | 287 | synthetic news articles (incl. distractors) |
|
| 34 |
-
| `
|
|
|
|
| 35 |
| `split.json` | — | `train` (207) / `test` (75) article split + seed |
|
| 36 |
-
| `unions.json` | — |
|
| 37 |
-
|
| 38 |
-
##
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
signed
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
(
|
| 47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 48 |
|
| 49 |
## Why synthetic
|
| 50 |
|
|
|
|
| 30 |
|
| 31 |
| File | Rows | Description |
|
| 32 |
|---|---|---|
|
| 33 |
+
| `articles.parquet` | 287 | input — synthetic news articles (incl. distractors) |
|
| 34 |
+
| **`nodes.parquet`** | **1,579** | **graph nodes** — actors per article (the entities) |
|
| 35 |
+
| **`gold.parquet`** | **914** | **graph edges** — gold signed relations between actors |
|
| 36 |
| `split.json` | — | `train` (207) / `test` (75) article split + seed |
|
| 37 |
+
| `unions.json` | — | raw nested form of the nodes (alias → canonical) |
|
| 38 |
+
|
| 39 |
+
## The signed graph (the core)
|
| 40 |
+
|
| 41 |
+
Each article has a **gold signed graph**: `nodes.parquet` (actors) + `gold.parquet`
|
| 42 |
+
(signed edges between them). Join them on (`article_id`, `union_id` ↔ `u_from`/`u_to`).
|
| 43 |
+
|
| 44 |
+
### `nodes.parquet` schema (the actors)
|
| 45 |
+
`article_id`, **`union_id`** (e.g. `U7`), **`type`** (`roster_actor` = known political
|
| 46 |
+
figure / `institutional_actor` / `non_roster_actor`), **`canonical_name`**, `surfaces`
|
| 47 |
+
(alias list). 1,579 nodes total.
|
| 48 |
+
|
| 49 |
+
### `gold.parquet` schema (the signed edges)
|
| 50 |
+
`article_id`, **`u_from`**, **`u_to`** (→ `nodes.union_id`), `act_type` (e.g. `endorses`,
|
| 51 |
+
`attacks`, `allies_with`, `questions`…), **`polarity`** (`positive` / `negative` /
|
| 52 |
+
`neutral` — **the edge sign**), `is_reactive`, `issue`, `evidence_quote`, `source`
|
| 53 |
+
(`planted` / …), `n_inclusion_votes`, `dispute_type`.
|
| 54 |
+
|
| 55 |
+
### `articles.parquet` schema (the input)
|
| 56 |
+
`article_id`, `title`, `body`, `dominio` (topic), **`dureza`** (relation difficulty:
|
| 57 |
+
`explicita` / `mixta` / `oblicua`), **`es_distractor`** (distractor flag).
|
| 58 |
|
| 59 |
## Why synthetic
|
| 60 |
|
articles.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b23e18a9f7af32a1ed0ba1a346fc15955dc5cf19330924a1163e3ececc45b352
|
| 3 |
+
size 355391
|
nodes.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a878e6a566716448c9f398b98cbd4b13d1969c22b7801bedfb7e1041572b429
|
| 3 |
+
size 26683
|