Pitch: 📎 GovHack 23.03_V1.pdf
The "Restruct Feedback" Hackathon challenge addresses a critical issue that arises during the draft legislation consultation process - unstructured text responses. Stakeholders are typically invited to provide feedback on draft legislation, which can result in an overwhelming number of responses that are disorganized and difficult to analyze. This challenge aims to streamline the process by creating a workflow that efficiently and effectively structures and categorizes feedback. This, in turn, will help the federal administration to understand and act on feedback quickly, reducing the time-consuming efforts involved in analyzing unstructured responses.
To achieve this, the proposed workflow involves using the ChatGPT API to analyze the sentiment of each response and categorize them based on the organization and summary of the main points. ChatGPT is a natural language processing tool developed by OpenAI that can understand and analyze language at an advanced level. By leveraging its capabilities, participants in the challenge can create a system that can effectively analyze and categorize unstructured text responses, making it easier for the federal administration to identify and address critical issues in the draft legislation.
Overall, the "Restruct Feedback" Hackathon challenge aims to simplify and streamline the draft legislation consultation process. By creating an efficient workflow that structures and categorizes unstructured text responses using the ChatGPT API, the federal administration can quickly understand and act on feedback, leading to more effective legislation development.
Organization: BK Federal Chancellery & Others
govtechhack23-RestructFeedback
Resources
Domain
- The sematic web in Fedlex https://fedlex.data.admin.ch/en-CH/home/intro
- Vernehmlassungsverordnung, VlV https://fedlex.data.admin.ch/eli/cc/2005/543
OpenAI
- Chatbot: https://chat.openai.com/chat
- OpenAI Cookbook -> must read: https://github.com/openai/openai-cookbook
- Playground -> test prompts against different models: https://platform.openai.com/playground
- Chat completion
- Guide: https://platform.openai.com/docs/guides/chat
- API reference (including parameters): https://platform.openai.com/docs/api-reference/chat/create>
- Text completion
- Guide: https://platform.openai.com/docs/guides/text
- API reference (including parameters): https://platform.openai.com/docs/api-reference/text/create>
- Models with limitations: https://platform.openai.com/docs/models/overview
- Model index for researchers: https://platform.openai.com/docs/model-index-for-researchers
- Compare output of different models: https://gpttools.com/comparisontool
- Frequency and presence penalties https://platform.openai.com/docs/api-reference/parameter-details
Previous
GovTech Hackathon 2023
Next project