Upload README.md with huggingface_hub

README.md

@@ -1,267 +1,162 @@
 ---
-dataset_info:
-- config_name: category_knn
-  features:
-  - name: src_category_id
-    dtype: string
-  - name: dst_category_id
-    dtype: string
-  - name: type
-    dtype: string
-  - name: weight
-    dtype: float32
-  splits:
-  - name: train
-    num_bytes: 55723
-    num_examples: 1248
-  download_size: 11432
-  dataset_size: 55723
-- config_name: category_nodes
-  features:
-  - name: category_idx
-    dtype: int64
-  - name: category_id
-    dtype: string
-  - name: name
-    dtype: string
-  - name: paper_count
-    dtype: int64
-  - name: x
-    dtype: float32
-  - name: y
-    dtype: float32
-  - name: z
-    dtype: float32
-  splits:
-  - name: train
-    num_bytes: 8216
-    num_examples: 156
-  download_size: 9128
-  dataset_size: 8216
-- config_name: paper_category_edges
-  features:
-  - name: src_paper_idx
-    dtype: int64
-  - name: dst_category_id
-    dtype: string
-  - name: type
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 63101007
-    num_examples: 1744492
-  download_size: 10161323
-  dataset_size: 63101007
-- config_name: paper_embeddings
-  features:
-  - name: paper_idx
-    dtype: int64
-  - name: paper_id
-    dtype: string
-  - name: canonical_paper_id
-    dtype: string
-  - name: embedding
-    sequence: float32
-  splits:
-  - name: train
-    num_bytes: 1577285149
-    num_examples: 1000000
-  download_size: 2179686943
-  dataset_size: 1577285149
-- config_name: paper_fulltext_embeddings
-  features:
-  - name: paper_idx
-    dtype: int64
-  - name: paper_id
-    dtype: string
-  - name: canonical_paper_id
-    dtype: string
-  - name: embedding
-    sequence: float32
-  splits:
-  - name: train
-    num_bytes: 1577285149
-    num_examples: 1000000
-  download_size: 2179663208
-  dataset_size: 1577285149
-- config_name: paper_knn
-  features:
-  - name: src_paper_idx
-    dtype: int64
-  - name: dst_paper_idx
-    dtype: int64
-  - name: type
-    dtype: string
-  - name: weight
-    dtype: float32
-  splits:
-  - name: train
-    num_bytes: 660002046
-    num_examples: 20000062
-  download_size: 250284912
-  dataset_size: 660002046
-- config_name: paper_nodes
-  features:
-  - name: paper_idx
-    dtype: int64
-  - name: paper_id
-    dtype: string
-  - name: canonical_paper_id
-    dtype: string
-  - name: paper_version
-    dtype: string
-  - name: title
-    dtype: string
-  - name: authors
-    dtype: string
-  - name: categories
-    sequence: string
-  - name: primary_category
-    dtype: string
-  - name: license
-    dtype: string
-  - name: update_date
-    dtype: string
-  - name: year
-    dtype: int32
-  - name: text_char_count
-    dtype: int64
-  - name: page_count
-    dtype: int64
-  - name: token_count
-    dtype: int64
-  - name: text_source
-    dtype: string
-  - name: text_is_partial
-    dtype: bool
-  - name: metadata_found
-    dtype: bool
-  - name: pdf_path
-    dtype: string
-  - name: x
-    dtype: float32
-  - name: y
-    dtype: float32
-  - name: z
-    dtype: float32
-  splits:
-  - name: train
-    num_bytes: 419637347
-    num_examples: 1000000
-  download_size: 181723940
-  dataset_size: 419637347
-- config_name: paper_topic_edges
-  features:
-  - name: src_paper_idx
-    dtype: int64
-  - name: dst_topic_id
-    dtype: string
-  - name: type
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 129793835
-    num_examples: 3000000
-  download_size: 52728344
-  dataset_size: 129793835
-- config_name: paper_year_edges
-  features:
-  - name: src_paper_idx
-    dtype: int64
-  - name: dst_year
-    dtype: int32
-  - name: type
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 24000000
-    num_examples: 1000000
-  download_size: 5823766
-  dataset_size: 24000000
-- config_name: topic_nodes
-  features:
-  - name: topic_idx
-    dtype: int64
-  - name: topic_id
-    dtype: string
-  - name: name
-    dtype: string
-  - name: paper_count
-    dtype: int64
-  - name: x
-    dtype: float32
-  - name: y
-    dtype: float32
-  - name: z
-    dtype: float32
-  splits:
-  - name: train
-    num_bytes: 110758080
-    num_examples: 1483957
-  download_size: 68132791
-  dataset_size: 110758080
-- config_name: year_nodes
-  features:
-  - name: year
-    dtype: int32
-  - name: paper_count
-    dtype: int64
-  - name: x
-    dtype: float32
-  - name: y
-    dtype: float32
-  - name: z
-    dtype: float32
-  splits:
-  - name: train
-    num_bytes: 456
-    num_examples: 19
-  download_size: 2584
-  dataset_size: 456
+pretty_name: Paper Universe Graph
+viewer: true
+tags:
+- datasets
+- graph
+- scientific-papers
+- arxiv
+- retrieval
+- embeddings
+size_categories:
+- 10M<n<100M
 configs:
-- config_name: category_knn
-  data_files:
-  - split: train
-    path: category_knn/train-*
-- config_name: category_nodes
+- config_name: paper_nodes
+  default: true
   data_files:
   - split: train
-    path: category_nodes/train-*
+    path: "parquet/paper_nodes.parquet"
 - config_name: paper_category_edges
+  default: false
   data_files:
   - split: train
-    path: paper_category_edges/train-*
-- config_name: paper_embeddings
+    path: "parquet/paper_category_edges.parquet"
+- config_name: paper_knn
+  default: false
   data_files:
   - split: train
-    path: paper_embeddings/train-*
-- config_name: paper_fulltext_embeddings
+    path: "parquet/paper_knn.parquet"
+- config_name: category_nodes
+  default: false
   data_files:
   - split: train
-    path: paper_fulltext_embeddings/train-*
-- config_name: paper_knn
+    path: "parquet/category_nodes.parquet"
+- config_name: category_knn
+  default: false
   data_files:
   - split: train
-    path: paper_knn/train-*
-- config_name: paper_nodes
+    path: "parquet/category_knn.parquet"
+- config_name: topic_nodes
+  default: false
   data_files:
   - split: train
-    path: paper_nodes/train-*
-  default: true
+    path: "parquet/topic_nodes.parquet"
 - config_name: paper_topic_edges
+  default: false
+  data_files:
+  - split: train
+    path: "parquet/paper_topic_edges.parquet"
+- config_name: year_nodes
+  default: false
   data_files:
   - split: train
-    path: paper_topic_edges/train-*
+    path: "parquet/year_nodes.parquet"
 - config_name: paper_year_edges
+  default: false
   data_files:
   - split: train
-    path: paper_year_edges/train-*
-- config_name: topic_nodes
+    path: "parquet/paper_year_edges.parquet"
+- config_name: paper_embeddings
+  default: false
   data_files:
   - split: train
-    path: topic_nodes/train-*
-- config_name: year_nodes
+    path: "parquet/paper_embeddings.parquet"
+- config_name: paper_fulltext_embeddings
+  default: false
   data_files:
   - split: train
-    path: year_nodes/train-*
+    path: "parquet/paper_fulltext_embeddings.parquet"
 ---
+
+# Paper Universe Graph Dataset
+
+Parquet-first export of the paper-universe graph already built under the Repository Library.
+
+This dataset preserves:
+
+- paper nodes with metadata references and 3D coordinates
+- paper/category/year/topic graph layers
+- optional paper-to-paper and category-to-category similarity edges
+- metadata and full-text paper embedding splits
+
+## Coverage
+
+- papers: `1000000`
+- categories: `156`
+- years: `19`
+- topics: `1483957`
+- embedding dimension: `384`
+- full-text embeddings included: `true`
+
+## Configs
+
+- default viewer/config: `paper_nodes`
+- `paper_nodes`: `1000000` rows
+- `paper_category_edges`: `1744492` rows
+- `paper_knn`: `20000062` rows
+- `category_nodes`: `156` rows
+- `category_knn`: `1248` rows
+- `topic_nodes`: `1483957` rows
+- `paper_topic_edges`: `3000000` rows
+- `year_nodes`: `19` rows
+- `paper_year_edges`: `1000000` rows
+- `paper_embeddings`: `1000000` rows
+- `paper_fulltext_embeddings`: `1000000` rows
+
+## Loading
+
+```python
+from datasets import load_dataset
+
+paper_nodes = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "paper_nodes")
+paper_category_edges = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "paper_category_edges")
+paper_knn = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "paper_knn")
+category_nodes = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "category_nodes")
+category_knn = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "category_knn")
+topic_nodes = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "topic_nodes")
+paper_topic_edges = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "paper_topic_edges")
+year_nodes = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "year_nodes")
+paper_year_edges = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "paper_year_edges")
+paper_embeddings = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "paper_embeddings")
+paper_fulltext_embeddings = load_dataset("YOUR_NAMESPACE/YOUR_DATASET", "paper_fulltext_embeddings")
+```
+
+## Notes
+
+- `paper_nodes` stores metadata references and coordinates, not the full paper body.
+- The original full text remains in the source paper dataset referenced by the manifest.
+- `paper_embeddings` is the metadata/title+abstract embedding split.
+- `paper_fulltext_embeddings` is the aggregated full-body embedding split when available.
+
+## Visualization Assets
+
+The export includes the local paper-universe visualizations and viewer payload when present:
+
+- `manifest.json`
+- `progress.json`
+- `render_manifest.json`
+- `viewer_manifest.json`
+- `universe_3d.png`
+- `universe_3d_detailed.png`
+- `nodes_3d_sample.html`
+- `universe_3d_hover.html`
+- `interactive/categories.json`
+- `interactive/manifest.json`
+- `interactive/papers_200000.json`
+- `interactive/papers_50000.json`
+- `interactive/years.json`
+
+### Universe 3D Overview
+
+![Paper universe 3D overview](./universe_3d.png)
+
+### Universe 3D Detailed View
+
+![Paper universe 3D detailed view](./universe_3d_detailed.png)
+
+- [Open the sampled 3D HTML view](./nodes_3d_sample.html)
+
+- [Open the interactive hover view](./universe_3d_hover.html)
+
+- Interactive viewer payload is included under `interactive/`.
+