chuyue commited on
Commit
a562675
·
verified ·
1 Parent(s): 416ed05

Remove unsourced rule grouping table

Browse files
Files changed (1) hide show
  1. README.md +13 -8
README.md CHANGED
@@ -78,16 +78,21 @@ Each Common Crawl dump is processed independently:
78
  7. **Duplicate removal:** remove duplicate followers using the generated blacklist.
79
  8. **Text sanitation:** repair residual HTML entities and mojibake, remove invisible characters, and reject corrupted or abnormally sized documents.
80
 
81
- The heuristic rules are organized into four groups:
82
 
83
- | Rule group | Main checks |
84
- |:---|:---|
85
- | Gopher Repetition | Duplicate lines and paragraphs, dominant n-grams, repeated-character ratios |
86
- | Gopher Quality | Document length, mean word length, symbol ratios, stop-word and alphabetic-word coverage |
87
- | C4 | Punctuation, boilerplate, JavaScript fragments, `lorem ipsum`, and policy text |
88
- | FineWeb Quality | Short lines, line-ending punctuation, repeated-line characters, and newline density |
89
 
90
- Deduplication is performed within each Common Crawl dump rather than across the full collection.
 
 
 
 
 
 
 
 
 
 
91
 
92
  ## 📁 Dataset Structure
93
 
 
78
  7. **Duplicate removal:** remove duplicate followers using the generated blacklist.
79
  8. **Text sanitation:** repair residual HTML entities and mojibake, remove invisible characters, and reject corrupted or abnormally sized documents.
80
 
81
+ Deduplication is performed within each Common Crawl dump rather than across the full collection.
82
 
83
+ ## 📈 Comparison with FineWeb
 
 
 
 
 
84
 
85
+ As a reproducibility reference, we compare URL sets on `CC-MAIN-2025-26`. After MinHash deduplication, the Ultra-FineWeb-L1 output shares 175,281,630 URLs with FineWeb. This is a pipeline comparison for one validation dump, not a quality ranking or a requirement that the two datasets contain the same documents.
86
+
87
+ | Metric | Before MinHash | After MinHash |
88
+ |:---|---:|---:|
89
+ | Ultra-FineWeb-L1 unique URLs | 258,071,345 | 194,402,835 |
90
+ | FineWeb unique URLs | 193,438,138 | 193,438,138 |
91
+ | URL intersection | 182,421,184 | 175,281,630 |
92
+ | Recall against FineWeb | 94.30% | **90.61%** |
93
+ | Precision | 70.69% | **90.16%** |
94
+
95
+ After MinHash deduplication, precision increases while recall decreases. This reflects a trade-off in URL coverage and deduplication; it is not a direct measurement of overall text quality. These figures apply only to the `CC-MAIN-2025-26` validation run and are not general claims about all released dumps.
96
 
97
  ## 📁 Dataset Structure
98