United KingdomPublic sector2024 – ongoingOfsted
GPT-4 Turbo analysis of Big Listen consultation responses
Ofsted analysed 27 consultation questions with GPT-4 Turbo and researchers validated every generated topic
Problem
Ofsted needed to identify recurring themes in written responses to its Big Listen consultation. Free-text answers required analysis before findings could be published.
Approach
Ofsted used GPT-4 Turbo through Azure OpenAI to propose recurring topics. It also used K-means clustering to find additional themes. Researchers checked generated descriptions against related respondent sentences and discarded topics that failed validation.
Results
Takeaway
Human validation determined which generated topics entered the findings. Ofsted did not publish a measured time saving for this work.
Categories
- Industry
- Public sector
- Business function
- Internal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AIPredictive analytics and ML
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Company's own claim
Sources
Source summaries
Ofsted documented AI-assisted analysis of responses to 27 free-text questions. Its methods statement described researcher validation of every generated topic before inclusion in the findings.
Full details are in the original sources.