--- dataset_info: - config_name: ar features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 333212298.1423473 num_examples: 829000 download_size: 177801306 dataset_size: 333212298.1423473 - config_name: az features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 5249155.0 num_examples: 18357 download_size: 2982366 dataset_size: 5249155.0 - config_name: bg features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 79264300.0 num_examples: 134795 download_size: 39006483 dataset_size: 79264300.0 - config_name: bn features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 30504148.0 num_examples: 50329 download_size: 12726950 dataset_size: 30504148.0 - config_name: ca features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 17343544.664760705 num_examples: 57740 download_size: 9269248 dataset_size: 17343544.664760705 - config_name: cs features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 113582742.32680418 num_examples: 392826 download_size: 74773901 dataset_size: 113582742.32680418 - config_name: da features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 191705980.58929488 num_examples: 609387 download_size: 117276590 dataset_size: 191705980.58929488 - config_name: de features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 236575466.97305882 num_examples: 798319 download_size: 133689762 dataset_size: 236575466.97305882 - config_name: el features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 208667801.70407987 num_examples: 349025 download_size: 91707457 dataset_size: 208667801.70407987 - config_name: en features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 11720097445.7493 num_examples: 40817655 download_size: 7854799132 dataset_size: 11720097445.7493 - config_name: es features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 274876034.1811792 num_examples: 797885 download_size: 168023555 dataset_size: 274876034.1811792 - config_name: et features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 16661155.0 num_examples: 56385 download_size: 9928608 dataset_size: 16661155.0 - config_name: fa features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 372259230.631925 num_examples: 841712 download_size: 184804243 dataset_size: 372259230.631925 - config_name: fi features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 114457190.20471013 num_examples: 373519 download_size: 70226166 dataset_size: 114457190.20471013 - config_name: fr features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 284087572.47052157 num_examples: 805753 download_size: 172392879 dataset_size: 284087572.47052157 - config_name: he features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 75530461.37165771 num_examples: 188530 download_size: 40129523 dataset_size: 75530461.37165771 - config_name: hi features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 228586526.43469945 num_examples: 478818 download_size: 98387329 dataset_size: 228586526.43469945 - config_name: hu features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 88644788.78502548 num_examples: 283635 download_size: 53164482 dataset_size: 88644788.78502548 - config_name: id features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 136738993.50897405 num_examples: 482649 download_size: 78061747 dataset_size: 136738993.50897405 - config_name: is features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 17543498.0 num_examples: 64358 download_size: 6622618 dataset_size: 17543498.0 - config_name: it features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 249171045.25293756 num_examples: 796614 download_size: 156693713 dataset_size: 249171045.25293756 - config_name: ja features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 278673154.2433263 num_examples: 662302 download_size: 169291425 dataset_size: 278673154.2433263 - config_name: ko features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 220764548.6417357 num_examples: 592344 download_size: 128210156 dataset_size: 220764548.6417357 - config_name: lt features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 28194308.833763536 num_examples: 89490 download_size: 17063680 dataset_size: 28194308.833763536 - config_name: lv features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 18552162.0 num_examples: 59446 download_size: 10706788 dataset_size: 18552162.0 - config_name: mr features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 12691415.0 num_examples: 26642 download_size: 5356253 dataset_size: 12691415.0 - config_name: ms features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 4490651.0 num_examples: 20867 download_size: 2171956 dataset_size: 4490651.0 - config_name: nl features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 275993909.47796786 num_examples: 802485 download_size: 172563957 dataset_size: 275993909.47796786 - config_name: 'no' features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 78516629.34054616 num_examples: 248062 download_size: 46230575 dataset_size: 78516629.34054616 - config_name: pl features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 255515110.77889696 num_examples: 782624 download_size: 166347323 dataset_size: 255515110.77889696 - config_name: pt features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 258642231.01709044 num_examples: 793637 download_size: 162908260 dataset_size: 258642231.01709044 - config_name: ro features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 112650093.33786418 num_examples: 378587 download_size: 63013817 dataset_size: 112650093.33786418 - config_name: ru features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 409291756.4775127 num_examples: 784173 download_size: 208916232 dataset_size: 409291756.4775127 - config_name: sk features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 42849959.97385041 num_examples: 141875 download_size: 27149894 dataset_size: 42849959.97385041 - config_name: sl features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 15414732.297041954 num_examples: 42019 download_size: 9838650 dataset_size: 15414732.297041954 - config_name: sv features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 204567446.80549085 num_examples: 675465 download_size: 128583128 dataset_size: 204567446.80549085 - config_name: ta features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 9489767.984177602 num_examples: 16305 download_size: 3511374 dataset_size: 9489767.984177602 - config_name: te features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 10803060.0 num_examples: 19578 download_size: 4210683 dataset_size: 10803060.0 - config_name: th features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 207397925.07035413 num_examples: 355047 download_size: 82636543 dataset_size: 207397925.07035413 - config_name: tl features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 5392722.0 num_examples: 16743 download_size: 2883565 dataset_size: 5392722.0 - config_name: tr features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 216709438.64376307 num_examples: 820040 download_size: 131050449 dataset_size: 216709438.64376307 - config_name: uk features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 255944319.26300222 num_examples: 580032 download_size: 134545628 dataset_size: 255944319.26300222 - config_name: vi features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 275315243.34182924 num_examples: 723073 download_size: 153396016 dataset_size: 275315243.34182924 - config_name: zh features: - name: query dtype: string - name: document dtype: string splits: - name: train num_bytes: 207574951.07480538 num_examples: 682722 download_size: 145982610 dataset_size: 207574951.07480538 configs: - config_name: ar data_files: - split: train path: ar/train-* - config_name: az data_files: - split: train path: az/train-* - config_name: bg data_files: - split: train path: bg/train-* - config_name: bn data_files: - split: train path: bn/train-* - config_name: ca data_files: - split: train path: ca/train-* - config_name: cs data_files: - split: train path: cs/train-* - config_name: da data_files: - split: train path: da/train-* - config_name: de data_files: - split: train path: de/train-* - config_name: el data_files: - split: train path: el/train-* - config_name: en data_files: - split: train path: en/train-* - config_name: es data_files: - split: train path: es/train-* - config_name: et data_files: - split: train path: et/train-* - config_name: fa data_files: - split: train path: fa/train-* - config_name: fi data_files: - split: train path: fi/train-* - config_name: fr data_files: - split: train path: fr/train-* - config_name: he data_files: - split: train path: he/train-* - config_name: hi data_files: - split: train path: hi/train-* - config_name: hu data_files: - split: train path: hu/train-* - config_name: id data_files: - split: train path: id/train-* - config_name: is data_files: - split: train path: is/train-* - config_name: it data_files: - split: train path: it/train-* - config_name: ja data_files: - split: train path: ja/train-* - config_name: ko data_files: - split: train path: ko/train-* - config_name: lt data_files: - split: train path: lt/train-* - config_name: lv data_files: - split: train path: lv/train-* - config_name: mr data_files: - split: train path: mr/train-* - config_name: ms data_files: - split: train path: ms/train-* - config_name: nl data_files: - split: train path: nl/train-* - config_name: 'no' data_files: - split: train path: no/train-* - config_name: pl data_files: - split: train path: pl/train-* - config_name: pt data_files: - split: train path: pt/train-* - config_name: ro data_files: - split: train path: ro/train-* - config_name: ru data_files: - split: train path: ru/train-* - config_name: sk data_files: - split: train path: sk/train-* - config_name: sl data_files: - split: train path: sl/train-* - config_name: sv data_files: - split: train path: sv/train-* - config_name: ta data_files: - split: train path: ta/train-* - config_name: te data_files: - split: train path: te/train-* - config_name: th data_files: - split: train path: th/train-* - config_name: tl data_files: - split: train path: tl/train-* - config_name: tr data_files: - split: train path: tr/train-* - config_name: uk data_files: - split: train path: uk/train-* - config_name: vi data_files: - split: train path: vi/train-* - config_name: zh data_files: - split: train path: zh/train-* --- # webfaq_cleaned A cleaned, multilingual question–document dataset across 44 languages, ~58.5M Q–D pairs total, intended as a contrastive training / retrieval-eval corpus. Derived from [`PaDaS-Lab/webfaq`](https://huggingface.co/datasets/PaDaS-Lab/webfaq). The source data is web-scraped FAQ pages. The raw input contained large amounts of SEO doorway pages, casino spam, e-commerce listing pages whose review snippet didn't match the product query, near-duplicate templates, and broken encodings. This dataset applies a multi-layer cleaning pipeline to filter that noise while preserving the substantive Q&A. ## Schema Each row is a single Q–D pair: | field | type | description | |----------|--------|----------------------------------------| | query | string | the question | | document | string | the answer / supporting passage | One config per language; one `train` split per config. ## Usage ```python from datasets import load_dataset # Load one language ds = load_dataset("bowang0911/webfaq_cleaned", "en", split="train") print(ds[0]) # {'query': '...', 'document': '...'} # Or stream it ds = load_dataset("bowang0911/webfaq_cleaned", "ja", split="train", streaming=True) for row in ds.take(5): print(row) ``` ## Languages 44 languages, sorted by final row count. `source` = rows pulled from the original `PaDaS-Lab/webfaq` shards; `kept` = rows remaining after cleaning. | lang | source | kept | kept% | lang | source | kept | kept% | |---|---:|---:|---:|---|---:|---:|---:| | en | 49,275,346 | 40,817,655 | 82.8% | ro | 546,709 | 378,587 | 69.2% | | fa | 969,748 | 841,712 | 86.8% | fi | 641,009 | 373,519 | 58.3% | | ar | 1,000,000 | 829,000 | 82.9% | th | 502,215 | 355,047 | 70.7% | | tr | 1,000,000 | 820,040 | 82.0% | el | 494,019 | 349,025 | 70.7% | | fr | 1,000,000 | 805,753 | 80.6% | hu | 363,866 | 283,635 | 78.0% | | nl | 1,000,000 | 802,485 | 80.2% | no | 316,320 | 248,062 | 78.4% | | de | 1,000,000 | 798,319 | 79.8% | he | 391,084 | 188,530 | 48.2% | | es | 1,000,000 | 797,885 | 79.8% | sk | 177,748 | 141,875 | 79.8% | | it | 1,000,000 | 796,614 | 79.7% | bg | 164,015 | 134,795 | 82.2% | | pt | 1,000,000 | 793,637 | 79.4% | lt | 119,192 | 89,490 | 75.1% | | ru | 1,000,000 | 784,173 | 78.4% | is | 91,630 | 64,358 | 70.2% | | pl | 1,000,000 | 782,624 | 78.3% | lv | 79,443 | 59,446 | 74.8% | | vi | 1,000,000 | 723,073 | 72.3% | ca | 80,189 | 57,740 | 72.0% | | zh | 1,000,000 | 682,722 | 68.3% | et | 69,643 | 56,385 | 81.0% | | sv | 832,140 | 675,465 | 81.2% | bn | 57,943 | 50,329 | 86.9% | | ja | 1,000,000 | 662,302 | 66.2% | sl | 49,867 | 42,019 | 84.3% | | da | 768,688 | 609,387 | 79.3% | mr | 30,658 | 26,642 | 86.9% | | ko | 646,996 | 592,344 | 91.6% | ms | 24,448 | 20,867 | 85.4% | | uk | 805,505 | 580,032 | 72.0% | te | 23,949 | 19,578 | 81.7% | | id | 652,660 | 482,649 | 74.0% | az | 19,738 | 18,357 | 93.0% | | hi | 537,488 | 478,818 | 89.1% | tl | 21,467 | 16,743 | 78.0% | | cs | 538,343 | 392,826 | 73.0% | ta | 19,815 | 16,305 | 82.3% | ## Cleaning Filters were developed iteratively: sample random rows per language, have an LLM inspect them to identify recurring failure patterns (templated SEO doorways, casino doorway pages, mismatched product-listing Q–Ds, encoding glitches), then encode those patterns as cheap substring/regex filters and re-sample to verify. Each row passes through four lightweight passes — a substring blocklist (cross-language SEO brand markers like KAYAK / momondo / Tripadvisor, plus a few per-language pattern sets added where sampling revealed language-specific noise ecosystems), a yes/no echo + query-template filter for the affected languages, near-duplicate dedup via MinHash LSH (64 perms, 16 bands, char-5gram shingles, Jaccard ≥ 0.85 per bucket), and encoding repair (HTML entities, JSON-unicode-escape leftovers, and literal `\r\n\t` fixed in place; rows still containing `\ufffd` after fix are dropped). The pipeline is deliberately conservative — it targets templated noise and known encoding glitches rather than scoring semantic relevance, so well-aligned Q–Ds are preserved even when the surface form is somewhat templated. ## Known limitations - **Languages where product-listing SEO is heavy** still have a residual ~15–20% mismatched pair rate after cleaning. The dominant failure mode is product-title queries paired with unrelated user reviews — each pair is unique (so dedup doesn't catch it) and lacks distinctive substring markers, so removing it cleanly requires a semantic Q–D relevance classifier rather than substring filtering. - **Languages with large kept% drops** (he, fi, ro, th, vi, el, cs, etc.) reflect heavy removal of templated travel-aggregator hotel-CTA pages, not loss of substantive QA. Spot checks on the kept rows for these languages show 90–100% substantive content. - **Other languages** (en, ar, hi, ko, de, fr, es, it, pt, ru, etc.) spot-check at 90–100% substantive. The remaining ~5–10% is mostly benign templated single-source answers (price predictions, weather summaries, hotel facts) rather than truly unaligned pairs. ## Source & citation This dataset is a derived, filtered version of [`PaDaS-Lab/webfaq`](https://huggingface.co/datasets/PaDaS-Lab/webfaq). If you use this dataset, please cite the original WebFAQ release as well as this cleaning pass. Intended for research use.