Entity Relations / Part of Speech Tagging
Named Entity Linking (PoS tagging) with the Universal Data Tool. Draw relationships between words or phrases within text.
Click "New File" on udt.dev
Then select the Text Entity Relations button from the
Setup > Data Type page.
Select Text Relations when choosing an interface
You can now configure the interface you'd like for you Text Entity Relations dataset by adding any labels you'd like to display per sample.
Text Entity Relation Configuration
You can use any of the following methods to import text data.
Go to the
Label tab to begin labeling data. See the Collaborative Labeling Guide to label with friends or a team of your labelers.
Example of Text Entity Relations labeling
The easiest way to use a Entity Relations dataset is using the JSON format. Use the "Download JSON" button at the top when you're done labeling and check out the Text Entity Relations JSON Specification.
Here's what a JSON sample looks like in the resultant dataset:
{
"document": "This strainer makes a great hat, I'll wear it while I serve spaghetti!",
"annotation": {
"entities": [
{
"text": "strainer",
"label": "hat",
"start": 5,
"end": 13,
"textId": "id1"
},
{
"text": "spaghetti",
"label": "food",
"start": 60,
"end": 69,
"textId": "id2"
},
{
"text": "I'll",
"start": 33,
"end": 37,
"textId": "id3"
},
{
"text": "wear",
"start": 38,
"end": 42,
"textId": "id4"
}
],
"relations": [
{
"from": "id3",
"to": "id4",
"label": "subject-doing"
}
]
}
}