OntExtract

Language Model Guided Document Analysis with Provenance Tracking

Research Preview: Under active development. Data may be reset, so download results regularly. Analytics via Umami (no cookies, no individual tracking). Email used for password reset only.

Research Workflow

Follow these steps to analyze semantic change in your documents

Define Terms

Create or import anchor terms to track their semantic evolution across time periods.

Upload Sources

Add background documents and scholarly works from different historical periods.

Create Experiment

Define your analysis experiment linking terms to document sets.

LLM Orchestration

AI analyzes your experiment and suggests optimal processing pipelines, or build your own manually.

Execute Pipeline

Review recommendations, adjust settings, and process your documents.

View Results

Explore semantic drift analysis, temporal patterns, and provenance graphs.

Research Access

Test document upload and processing.
Early beta - your feedback helps improve the system.