Athena News API Documentation
FeaturesPricingContact UsCreate AccountSign In
Athena API v3
Athena API v3
  • About Athena News API
  • v3 Change Summary
  • Account Overview
    • Account Overview
    • Dashboard
    • News Search
    • API Key
    • Account Settings
    • Manage Subscription
  • Getting Started
    • API Quickstart
    • Python SDK
    • Intro to Building a Query
    • API Reference
      • News Search
      • Headlines (BETA)
    • Data Model Reference
      • Articles
      • Entities
    • Supported Languages
    • Rate Limits
    • Troubleshooting
Powered by GitBook
On this page
  1. Getting Started
  2. Data Model Reference

Entities

PreviousArticlesNextSupported Languages

Last updated 1 month ago

Entity extract is broken down into three steps.

  1. Extraction: Entities, their associated actions and relationships to other entities are identified.

  2. Linkage: Entities are linked/resolved to Wikidata and other knowledge bases.

  3. Sentiment Analysis: Sentiment for the entity is calculated based on their associated actions.

    1. Sentiment Score: The score exists on a scale between -1 and 1. Anything below -0.5 should be considered negative, anything above +0.5 should be considered positive.

    2. Confidence Score: A confidence score indicates the relative level of confidence in that sentiment score. It is calculated based on the deviation between multiple actions and weighted by the number of actions that have been identified.