Datasets:
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
url: large_string
split: large_string
label: large_string
ru_score: double
ua_score: double
text_excerpt: large_string
-- schema metadata --
pandas: '{"index_columns": [], "column_indexes": [], "columns": [{"name":' + 761
to
{'pattern': Value('large_string'), 'direction': Value('large_string'), 'language': Value('large_string'), 'legal_source': Value('large_string'), 'weight': Value('float64')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
return get_rows(
^^^^^^^^^
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/utils.py", line 77, in get_rows
rows_plus_one = list(itertools.islice(ds, rows_max_number + 1))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2690, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2227, in __iter__
for key, pa_table in self._iter_arrow():
^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2251, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 494, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 384, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/parquet/parquet.py", line 209, in _generate_tables
yield Key(file_idx, batch_idx), self._cast_table(pa_table)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/parquet/parquet.py", line 147, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2321, in table_cast
return cast_table_to_schema(table, schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2249, in cast_table_to_schema
raise CastError(
datasets.table.CastError: Couldn't cast
url: large_string
split: large_string
label: large_string
ru_score: double
ua_score: double
text_excerpt: large_string
-- schema metadata --
pandas: '{"index_columns": [], "column_indexes": [], "columns": [{"name":' + 761
to
{'pattern': Value('large_string'), 'direction': Value('large_string'), 'language': Value('large_string'), 'legal_source': Value('large_string'), 'weight': Value('float64')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Crimea Digital Sovereignty: C4 Training Corpus Analysis
Summary statistics from a full census of sovereignty framing in Google's C4 training corpus. A Rust-based classifier with 90 legal-grounded signals scanned 34.1M Crimea-mentioning documents across English, Russian, and Ukrainian splits.
Description
This dataset contains summary-level results (not the full 891K classified documents). The classifier uses deterministic regex with no ML/learned parameters. Each of the 90 signals is grounded in legal provenance (OFAC SDN, EU Regulations, GEC reports, Russian federal law).
Full classifier source code: c4_sovereignty/scanner
Configs
| Config | Rows | Description |
|---|---|---|
summary |
4 | Per-split totals (EN, UK, RU, Total) |
source_tiers |
7 | Source classification across 7 tiers |
sample |
100 | Random classified excerpts (50 EN + 50 UK) |
Key Statistics
| Metric | Value |
|---|---|
| Total documents scanned | 34,132,798 |
| Russia-framing documents | 891,522 (2.61%) |
| Ukraine-framing documents | 1,271,984 |
| Classifier signals | 90 |
| From independent sources | 95.3% |
| From state/sanctioned sources | 4.7% |
Per-split breakdown
| Split | Total docs | Russia-framing | % |
|---|---|---|---|
| English | 286,117 | 3,607 | 1.26% |
| Ukrainian | 3,639,461 | 6,271 | 0.17% |
| Russian | 30,207,220 | 881,644 | 2.92% |
| Total | 34,132,798 | 891,522 | 2.61% |
Source classification of 891,522 Russia-framing documents
| Source tier | Count | % | Legal provenance |
|---|---|---|---|
| Independent | 849,761 | 95.3% | No state ties identified |
| State-adjacent | 16,772 | 1.9% | Sberbank/Gazprom, EU Pkg 16 |
| State media (T1) | 14,406 | 1.6% | OFAC SDN, EU Regulations |
| Sanctioned proxy | 4,928 | 0.6% | GEC 2020, OFAC EO14024 |
| Government | 3,156 | 0.4% | Russian federal law |
| State-controlled (T2) | 2,470 | 0.3% | Gazprom Media/NMG |
| Pravda network | 29 | <0.1% | VIGINUM/SGDSN 2024 |
Validation
Validated by two independent annotators on 100 random Russia-framing documents per language split (EN, UK, RU; 300 total). Blind review — no classifier scores visible.
- EN: kappa = 0.94, precision 89.6%
- UK: kappa = 0.56, precision 95.0%
- RU: kappa = 0.49, precision 97.0%
- Weighted average precision: 93.9% (278/296)
See: crimea-sovereignty-validation
Citation
@misc{dobrovolskyi2026digital,
author = {Dobrovolskyi, Ivan},
title = {Digital Annexation: A Computational Audit of Crimea's Sovereignty Framing in Large Language Models},
year = {2026},
url = {https://crimeaisukraine.org}
}
License
CC-BY-4.0
- Downloads last month
- 130