Update README.md

README.md

@@ -3261,3 +3261,31 @@ configs:
   - split: label
     path: zh-tw/label-*
 ---
+
+Entity label data extracted from Wikidata (wikidata-20220103-all.json.gz), filtered for item entities only  
+Similar dataset: https://huggingface.co/datasets/wmt/wikititles (Wikipedia titles instead of all Wikidata entities)
+
+You can generate parallel text examples from this dataset like below:
+
+```python
+from datasets import load_dataset
+import pandas as pd
+
+def get_lang_pair(lang_codes: list, how="inner", repo_id="rayliuca/wikidata_entity_label", merge_config={}, datasets_config={}) -> pd.DataFrame:
+    out_df = None
+    for lc in lang_codes:
+        dataset = load_dataset(repo_id, lc, **datasets_config)
+        dataset_df = dataset['label'].to_pandas().rename(columns={"label":lc}).drop(columns=['lastrevid'])
+        if out_df is None:
+            out_df = dataset_df
+        else:
+            out_df = out_df.merge(
+                    dataset_df,
+                    on='wikidata_id',
+                    how=how,
+                    **merge_config
+                )
+    return out_df
+
+get_lang_pair(['en', 'fr', 'ja', 'zh'])
+```