Update README.md

README.md

@@ -17,7 +17,9 @@ source_datasets:
 task_categories:
 - token-classification
 task_ids:
-- named-entity-recognition
+- token classification
+- named entity recognition
+- relation extraction
 ---
 
 # PET: A NEW DATASET FOR PROCESS EXTRACTION FROM TEXT
@@ -32,6 +34,8 @@ task_ids:
   - [Languages](#languages)
 - [Dataset Structure](#dataset-structure)
   - [Data Instances](#data-instances)
+    - [Token Classification](#data-instances-token-classification)
+    - [Relations Extraction](#data-instances-relations-extraction)
   - [Data Fields](#data-fields)
   - [Data Splits](#data-splits)
 - [Dataset Creation](#dataset-creation)
@@ -75,6 +79,8 @@ Test set to beanchmark *Business Process Extraction from Text* approaches.
 
 ### Data Instances
 
+#### <a name="data-instances-token-classification"></a>Token Classification
+
 For each instance, there is a document name representing the name of the document of the Friedrich *et al.* dataset, an integer representing the number of the sentence, a list of tokens representing the words of the sentence instance, and a list of *ner tags* (in IOB2 format) representing the annotation of process elements of the sentence.
 
 Below, an example of data instance.
@@ -87,6 +93,8 @@ Below, an example of data instance.
 }
 ```
 
+#### <a name="data-instances-relations-extraction"></a>Token Classification
+
 ### Data Fields
 
 - *document name*: a string used to represent the name of the document.