Audio-to-text parliamentary transcription system

Transcription and translation

Israel

Use case ID: 047

Author: Parliament of Israel

Date: 10 Jul 2024

Objective:

Transcribe recorded audio files into text, supporting various parliamentary functions, such as maintaining official records of parliamentary speeches and transcribing public hearings, thereby shortening the time required to produce protocols, and reducing hand-related medical issues (such as repetitive strain injury) among parliamentary stenographers.

Actors:

  • Stenographers
  • IT department
  • Data & information department
  • Protocols department

Prerequisites:

  • High-end recording system in the plenum
  • Artificial intelligence (AI) model trained for speech-to-text transcription
  • Integration with stenographers’ digital management system (E-Parliament)

Scenario:

  1. When a parliamentary meeting starts, the audio recordings are sent in real time to the automatic speech recognition (ASR) engine to produce a text transcript.
  2. Stenographers review and correct portions of the preliminary diarized transcript to produce the current segment of the parliamentary meeting for publication.
  3. When the parliamentary meeting ends, the system combines all the text transcripts into one protocol and publishes it on the parliamentary website.

Alternate flows:

  • The process is done manually.

Expected results:

  • Workload for stenographers is reduced.
  • Protocols are produced and distributed more quickly.
  • The process of transcribing speech is more efficient.
  • The quality of protocols is increased. 

Potential challenges:

  • Accuracy of ASR-produced protocols
  • Issues with the output of the ASR system caused by issued with recorded sound quality
  • AI model training based on transcription history, which might be incomplete or contain potential inaccuracies. 

Data requirements:

  • Historically recorded speech recordings for enhancing the performance of AI models
  • Periodic verification of the AI ​​model’s performance
  • Historic plenum speech protocols

Integrations with other systems:

  • Stenographers’ digital management system (E-Parliament)
  • Legislation system (Sanhedrin)
  • Public parliamentary website
  • AI-powered ASR system
  • Business intelligence reporting tools
  • Plenum audio/video recording system

Success metrics:

  • Automatic transcription process time
  • Word error rate (WER)
  • Multiple concurrent ASR processes with no performance degradation

 

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].