failed09/bashkir-frequency-index

Bashkir Frequency Index v11.0

Word-frequency index for the Bashkir language, computed over 136.41 million tokens of clean monolingual text across diverse public domains (periodicals, news media, literary publications, encyclopedic texts, books, and general web archives). Non-Bashkir language admixture and scanning artifacts were filtered using automated language-filtering pipelines.

Configurations

Config	Rows	Cutoff	Use case
public (recommended)	603,275	count ≥ 3	NLP applications, spellchecking, text processing, linguistic research
full	843,038	count ≥ 2	Comprehensive morphology studies, lexical research

Additionally provided in the repository:

• top_coverage_50.json: the minimal set of 1,160 word forms covering 50% of all tokens in the corpus.
• top_500.json: top-500 core frequency vocabulary list.

Schema

Column	Type	Description
rank	int64	Rank by descending frequency
word	string	Lowercase Cyrillic word form
count	int64	Absolute occurrence count
doc_freq	int64	Number of sentences containing the form
freq_per_million	double	Normalized frequency per million indexed tokens

Sample (Top 10 Words)

Rank	Word	Function / Translation	Count	FPM	Doc Frequency
1	һәм	conjunction "and"	2,104,941	15,277.5	1,642,308
2	менән	postposition "with"	1,602,535	11,631.1	1,324,248
3	бер	numeral "one / a certain"	1,012,859	7,351.3	832,451
4	ла	particle "also / and"	821,024	5,958.9	668,232
5	был	pronoun "this"	788,218	5,720.8	693,574
6	ул	pronoun "he / she / it / that"	752,627	5,462.5	662,875
7	өсөн	postposition "for / sake of"	745,924	5,413.9	636,682
8	тип	quotative gerund "saying / as"	709,534	5,149.7	616,259
9	лә	particle "also / and"	521,084	3,782.0	435,357
10	буйынса	postposition/link "according to"	462,733	3,358.5	409,618

Coverage

Tokens covered	Forms needed
25%	145
50%	1,160
80%	11,651
90%	34,798
95%	83,991

Method

• Tokenization: Regex lowercase tokenization over Cyrillic script including all 9 Bashkir-specific letters (Ә Ғ Ҙ Ҡ Ң Ө Ҫ Ү Һ).
• Corpus Base: 9.70M sentences (136.41M indexed tokens).
• Filtering: Automated function-word language filtering to remove foreign language leaks.
• Form Constraints: Isolated single-character forms removed (except ә); concatenated strings (>24 chars) removed.

Loading

from datasets import load_dataset
dataset = load_dataset("failed09/bashkir-frequency-index", "public")
df = dataset["train"].to_pandas()
print(df.head(10))

Quality and Use

Non-Bashkir words and other language admixture were partially removed using automatic filtering. Some non-Bashkir words, borrowed forms, names, technical terms, or noisy entries may still remain. Always validate final results before using this index in a production system, linguistic analysis, or user-facing application. With each new release, we will continue improving filtering, coverage, and overall quality.

Citation

@dataset{failed09_bashkir_frequency_index_2026,
  title     = {Bashkir Frequency Index v11.0},
  author    = {failed09},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/failed09/bashkir-frequency-index}
}

Open data for the preservation and development of the Bashkir language.