Skip to main content

Media API: Guide

The Geneea Media API is designed to automate metadata extraction, content categorization, and content discovery for publishers. Its features and configuration options are built to support two primary workflows:

  • Interactive use within a CMS: Offered as our Newsroom Assistant, these features are designed to be integrated directly into your CMS. This allows journalists to review and approve reader-facing tags, entities, and photo recommendations before publication.
  • Fully automated background processing: Tasks like assigning industry-standard taxonomies (e.g., IPTC Media Topics) for analytics, linking related articles, or processing historical archives require no manual supervision.

This guide serves as an entry point to understanding and implementing these capabilities.

Features​

Semantic Tags​

Semantic tags are the most important labels describing an article's overall content. They include relevant entities, broad topics, and industry-standard taxonomy categories (such as IPTC Media Topics and IAB Content Taxonomy).

Entities​

Entities provide granular details about exactly who or what is mentioned in an article. The API extracts explicitly named entities (people, locations, organizations, events), general concepts (keywords like flu season), and derived entities (concepts inferred geographically or logically).

  • Working with Entities: Detailed information on entity types, standard forms, mention extraction, and relevance scoring.

Related Content Recommendations​

Beyond metadata extraction, the API helps keep readers engaged by automatically suggesting relevant media and articles based on semantic similarity.

  • Related Articles: Recommend other articles from your database covering similar topics. This matching is driven by the underlying semantic tags and the relevance of extracted entities.
  • Photo Recommendations: Automatically suggest relevant images for an article based on its semantic content. Recommendations can be sourced from your licensed providers, our partners, or public domains like Wikipedia.

Technical Concepts & Integration​

The following guides cover the technical architecture, configuration options, and operational aspects of using the Media API.

Knowledge & Customization​

  • Geneea Knowledge Base (GKB): Understand how tags and entities are mapped to our central knowledge graph. The GKB integrates open data (like Wikidata) with proprietary identifiers, and supports custom properties unique to your organization.
  • Feedback-Based Learning: Learn how the API adapts to your specific editorial preferences. By passing editor corrections (confirming or rejecting tags) back to the API, the system continuously improves its accuracy for your specific use case.

Languages & Localization​

  • Supported Languages: An overview of the languages supported by our analysis models.
  • Presentation Language: Learn how to request API outputs (such as entity names and taxonomy labels) in a specific language, regardless of the article's original language. This is highly useful for multilingual publishing and centralized analytics.

Operations & Management​

  • Archive Analysis: Best practices for processing large historical datasets and back-catalogs of articles through the API.
  • Billing: Information on how API requests are counted, usage limits, and billing metrics.