Collage
A tool designed for rapid prototyping, visualization, and evaluation of different information extraction models on scientific PDFs
Collage is a tool designed for rapid prototyping, visualization, and evaluation of different information extraction models on scientific PDFs. Further, we enable both non-technical users and NLP practitioners to inspect, debug, and better understand modeling pipelines by providing granular views of intermediate states of processing.
You can find more information about Collage in the paper, which was published at the ACL 2025 Workshop on Scholarly Document Processing.
This demo should be available and running at this URL. This server can sometimes be unstable. If it is having issues when you try to access it, please follow the Docker Compose instructions on the GitHub Repo.