NER (named entity recognition) is a subtask where entities that appear in the content are identified and output with their determined category. ELSA (entity level sentiment analysis) is also part of the NER output. The relating fields (sentimentValue
, sentimentScore
) are included below value
in the response and provides the sentiment in which the specific entity was mentioned. The NER output value
is derived based on the following categories: organization
, person
, location
, or miscellaneous
.
Note: Clients must reach out to their Account Manager for access to NER and ELSA.
Syntax | Code Block |
---|
"ner": [
{
"value": "<Identified Entity>",
"sentimentValue": "<Positive, Neutral or Negative>",
"sentimentScore": <-1 - 1>
"type": " |
|
<IAB v3.0 category Name><organization, person, location, or miscellaneous>"
}
]
} |
|
Example | Code Block |
---|
"ner": [
{
"value": "vegas",
"sentimentValue": "neutral",
"sentimentScore": 0.296,
"type": "location"
},
{
"value": "us",
"sentimentValue": "neutral",
"sentimentScore": 0,
"type": "location"
},
{
"value": "fanatics",
"sentimentValue": "neutral",
"sentimentScore": 0,
"type": "organization"
},
{
"value": "patriots",
"sentimentValue": "neutral",
"sentimentScore": 0,
"type": "organization"
},
{
"value": "keisha epps",
"sentimentValue": "neutral",
"sentimentScore": 0,
"type": "person"
},
{
"value": "matt barnes",
"sentimentValue": "neutral",
"sentimentScore": 0,
"type": "person"
},
{
"value": "jumanji",
"sentimentValue": "negative",
"sentimentScore": -0.4767,
"type": "miscellaneous"
}
], |
|
Languages | All supported languages. |
Data Type | Array of objects |
More Information | IAB Tech Lab Content Taxonomy ID - V3