Add 'document_id' and 'sentence_id' columns

README.md

@@ -230,10 +230,13 @@ There are two languages available : Spanish (es) and Dutch (nl).
 The examples look like this :
 
 ```
-{'id': '0',
- 'ner_tags': [5, 6, 0, 0, 0, 0, 3, 0, 0],
- 'pos_tags': [4, 28, 13, 59, 28, 21, 29, 22, 20],
- 'tokens': ['La', 'Coruña', ',', '23', 'may', '(', 'EFECOM', ')', '.']
+{
+  'id': '0',
+  'document_id': 0,
+  'sentence_id': 0,
+  'tokens': ['Melbourne', '(', 'Australia', ')', ',', '25', 'may', '(', 'EFE', ')', '.'],
+  'pos_tags': [29, 21, 29, 22, 13, 59, 28, 21, 28, 22, 20],
+  'ner_tags': [5, 0, 5, 0, 0, 0, 0, 0, 3, 0, 0]
 }
 ```
 
@@ -244,6 +247,8 @@ Indeed `-DOCSTART-` is a special line that acts as a boundary between two differ
 ### Data Fields
 
 - `id`: id of the sample
+- `document_id`: an `int32` feature tracking which document the sample is from.
+- `sentence_id`: an `int32` feature tracking which sentence in this document the sample is from.
 - `tokens`: the tokens of the example text
 - `ner_tags`: the NER tags of each token
 - `pos_tags`: the POS tags of each token

conll2002.py

@@ -85,6 +85,8 @@ class Conll2002(datasets.GeneratorBasedBuilder):
             features=datasets.Features(
                 {
                     "id": datasets.Value("string"),
+                    "document_id": datasets.Value("int32"),
+                    "sentence_id": datasets.Value("int32"),
                     "tokens": datasets.Sequence(datasets.Value("string")),
                     "pos_tags": datasets.Sequence(
                         datasets.features.ClassLabel(
@@ -197,18 +199,26 @@ class Conll2002(datasets.GeneratorBasedBuilder):
         logger.info("⏳ Generating examples from = %s", filepath)
         with open(filepath, encoding="utf-8") as f:
             guid = 0
+            document_id = 0
+            sentence_id = 0
             tokens = []
             pos_tags = []
             ner_tags = []
             for line in f:
                 if line.startswith("-DOCSTART-") or line == "" or line == "\n":
+                    if line.startswith("-DOCSTART-"):
+                        document_id += 1
+                        sentence_id = 0
                     if tokens:
                         yield guid, {
                             "id": str(guid),
+                            "document_id": document_id,
+                            "sentence_id": sentence_id,
                             "tokens": tokens,
                             "pos_tags": pos_tags,
                             "ner_tags": ner_tags,
                         }
+                        sentence_id += 1
                         guid += 1
                         tokens = []
                         pos_tags = []
@@ -219,10 +229,13 @@ class Conll2002(datasets.GeneratorBasedBuilder):
                     tokens.append(splits[0])
                     pos_tags.append(splits[1])
                     ner_tags.append(splits[2].rstrip())
-            # last example
-            yield guid, {
-                "id": str(guid),
-                "tokens": tokens,
-                "pos_tags": pos_tags,
-                "ner_tags": ner_tags,
-            }
+            if tokens:
+                # last example
+                yield guid, {
+                    "id": str(guid),
+                    "document_id": document_id,
+                    "sentence_id": sentence_id,
+                    "tokens": tokens,
+                    "pos_tags": pos_tags,
+                    "ner_tags": ner_tags,
+                }