The "context" project aims to create an educational tool for simplifying the process of knowledge management.
-
Step 1: Collection
- Simple, mobile-friendly web-app with login.
- Data collection through submission of URLs, text, images, documents for async processing.
-
Step 2: Content Extraction
- Automatically extract text content from submitted URLs.
-
Step 3: Knowledge Connection
- Extract meta-data
- Extract entities, topics, ...
- Generate embeddings
- Create one or multiple knowledge graphs
-
Step 4: Knowledge Exploration
- Query the graph
- Visually inspect the graph
- Natural language interface (chat) with RAG
-
Step 5: Feedback Loop
- Enable users to provide feedback on auto-generated knowledge graph connections and embeddings for continuous improvement