Thematic classification of portal content

Classification systems

Brazil - Chamber of Deputies

Use case ID: 018

Author: Chamber of Deputies of Brazil

Date: 14 June 2024

Objective:

Read the text of bills, laws, news articles, speeches, studies and other content available on the portal, interpret this content, and classify it according to a fixed set of macro themes (e.g. economics, health, technology, human rights). This enables thematic pages to be automatically generated on the portal.

Actors:

  • Portal Content and Management staff
  • Information Management staff
  • AI system for thematic classification of portal content

Prerequisites:

  • Macro themes validated by both the Portal Content and Management staff and the Information Management staff
  • AI model trained for thematic classification
  • Integration with digital services used to supply the portal with new content, such as bills, laws, news articles and legislative studies
  • Seamless integration with a powerful search and analytics engine

Scenario:

  1. When a new bill (or other relevant content for the portal) is inserted into the appropriate digital service, it is indexed by the search engine and thematically classified by the AI system, without requiring any user intervention.

Alternate flows:

  • When content is incorrectly classified by the AI system, a Portal Content and Management staff member or an Information Management staff member can assign a new classification. This feedback should be utilized to retrain the AI model and enhance future classifications.

Expected results:

  • Portal users seeking access to content centered on specific macro themes, such as health, economics or technology, have their needs fulfilled.
  • Biases in the classification of thematic portal content are mitigated.
  • Workload for Portal Content and Management staff is reduced.

Potential challenges:

  • Ensuring the AI model accurately classifies content amid emerging new topics
  • Detecting potential deterioration in the accuracy of the AI system over time
  • Encouraging users to provide feedback

Data requirements:

  • Historically themed and classified content for training AI models
  • Bidirectional API-driven data exchange linking the search engine and the AI model
  • Periodic verification of AI ​​model accuracy

Integrations with other systems:

  • Digital services used to supply the portal with new content
  • Enhanced search and analytics engine with integrated OCR and advanced indexing features
  • Analytics and reporting tools

Success metrics:

  • Accuracy rate of content classification by the AI model (Top-1 accuracy)

 

The Use cases for AI in parliaments collection is published by the IPU’s Centre for Innovation in Parliament as part of the Parliamentary Data Science Hub’s project to create guidelines for AI governance in parliaments.

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International licence. It may be freely shared and reused with acknowledgement of the author and the IPU. 

A use case describes how a system should work. It is used to plan, develop and measure implementation. A use case is not the same as a case study, which is a descriptive text of an actual project’s implementation. Please note that this use case is provided “as is” and neither the IPU nor the author accepts any responsibility for its use.

For more information about the IPU’s work on artificial intelligence, please visit www.ipu.org/AI or contact [email protected].