Datasets:

LVSTCK
/

macedonian-corpus-cleaned

Modalities:

Formats:

Languages:

Size:

ArXiv:

Tags:

Libraries:

License:

Dataset card Data Studio Files Files and versions Community

StefanKrsteski commited on Jan 2

Commit

41a2a61

·

verified ·

1 Parent(s): 9bd9fdf

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -42,7 +42,7 @@ This implementation applies heuristic rules derived from the [C4 dataset](https:
   **5. PII Filtering.**
   - Removed all Personally Identifiable Information (PII), including: Email addresses, IP addresses, Phone numbers.
-  **6. Text Chunking and Cleaning:** Breaks texts into manageable chunks, each not exceeding 4000 characters, applied only for data sourced from MMORE.  This step also involves correcting common errors that were identified after qualitative evaluation, deleting specific unwanted patterns texts.
 As a further cleaning step, we performed MinHash Deduplication after step 6. The deduplicated dataset is available [here](https://huggingface.co/datasets/LVSTCK/macedonian-corpus-cleaned-dedup).

   **5. PII Filtering.**
   - Removed all Personally Identifiable Information (PII), including: Email addresses, IP addresses, Phone numbers.
+  **6. Text Chunking and Cleaning:** Breaks texts into manageable chunks, each not exceeding 4000 characters while respecting sentence boundaries, applied only for data sourced from MMORE. This step also involves correcting common errors that were identified after qualitative evaluation, deleting specific unwanted patterns texts.
 As a further cleaning step, we performed MinHash Deduplication after step 6. The deduplicated dataset is available [here](https://huggingface.co/datasets/LVSTCK/macedonian-corpus-cleaned-dedup).