changed notation scheme to IOB

README.md

@@ -3,7 +3,7 @@ annotations_creators:
 - other
 language_creators:
 - found
-languages:
+language:
 - ro
 license:
 - cc-by-nc-nd-4.0
@@ -89,6 +89,8 @@ The files contain the following data fields
     - `TIME`: Time reference
     - `O`: No entity annotation present
 
+The final tagset (in IOB notation) is the following: `['O', 'B-TIME', 'I-TIME', 'B-LEGAL', 'I-LEGAL', 'B-ORG', 'I-ORG', 'B-LOC', 'I-LOC', 'B-PER', 'I-PER']`
+
 ### Data Splits
 
 Splits created by Joel Niklaus. 

convert_to_hf_dataset.py

@@ -1,5 +1,4 @@
 import os
-import re
 from glob import glob
 from pathlib import Path
 
@@ -17,9 +16,6 @@ base_path = Path("legalnero-data")
 tokenizer = Romanian().tokenizer
 
 
-# A and D are different government gazettes
-# A is the general one, publishing standard legislation, and D is meant for legislation on urban planning and such things
-
 def process_document(ann_file: str, text_file: Path, metadata: dict, tokenizer) -> List[dict]:
     """Processes one document (.ann file and .txt file) and returns a list of annotated sentences"""
     # read the ann file into a df
@@ -55,7 +51,7 @@ def process_document(ann_file: str, text_file: Path, metadata: dict, tokenizer)
                 print(f"Could not find entity `{row['entity_text']}` in sentence `{sentence}`")
 
         ann_sent["words"] = [str(tok) for tok in doc]
-        ann_sent["ner"] = [tok.ent_type_ if tok.ent_type_ else "O" for tok in doc]
+        ann_sent["ner"] = [tok.ent_iob_ + "-" + tok.ent_type_ if tok.ent_type_ else "O" for tok in doc]
 
         annotated_sentences.append(ann_sent)
     if not_found_entities > 0:
@@ -89,6 +85,8 @@ df.ner = df.ner.apply(lambda x: x[:-1])
 # remove rows with containing only one word
 df = df[df.words.map(len) > 1]
 
+print(f"The final tagset (in IOB notation) is the following: `{list(df.ner.explode().unique())}`")
+
 # split by file_name
 num_fn = len(file_names)
 train_fn, validation_fn, test_fn = np.split(np.array(file_names), [int(.8 * num_fn), int(.9 * num_fn)])

test.jsonl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8009d0a864d6ffe4174cd876afe7bf5f5c01cfb31d1165b76a9f5d2eecd23b85
-size 409786
+oid sha256:3c73ba1c9e453754d9efb64a6b94db9ffd0997a36dd4d16a7fcccd449c266513
+size 414576

train.jsonl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0987ce4839662c5c4144aec5f9dbcb3669abb7143d45718e47af486b64cf0829
-size 3266615
+oid sha256:59f76a7b1c48596c8d3a6b0c19b7c4ff76ca991df091e0c3cb3447ee57386aa7
+size 3301849

validation.jsonl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:84b8948d38ef736d39849811dc7648b1086903bf7efabb86c6ac266175575d57
-size 421295
+oid sha256:5b81f444ea43588cb3a62968df0d6708aa131754fe7dffee1ba94b5e0757a307
+size 426257