Proof of Concept Chatbot
- 1 minute to read
Problem
Digital archives have traditionally required tagging so users can search accurately. How can we develop an easier way for archivists and the community to summarise their documents, and make it easier for their users to access what they want?
Solution
We are using a chatbot to allow searching for relevant documents using natural language. This product is called Archaiv. We will allow searching using natural language which will return both a natural language response and structured tags. A diverse and robust governance board will work with the community to reduce bias.
Chatbot
Our chatbot was built so users could easily search for digital objects that were important to them. Find the code to run our bot at https://github.com/archaivgovhack/archaivgovhack.github.io
Thanks to AWS for the instance hosting. Note, using OpenAI ChatGPT API - can be slow.
Experimental: If an error occurs, simply refresh the page.
Example prompts:
- What content do you have?
- What context are you aware of, what do you explicitly not know?
- What do you think is important?
- What is GovHack?