PaperMage: A Unified Toolkit for Processing, Representing, and Manipulating Visually-Rich Scientific Documents

Kyle Lo·Shannon Zejiang Shen·Benjamin Newman·Joseph Chee Chang
EMNLP·2023·24 citations🏆 Best Demo

AbstractDespite growing interest in applying natural language processing (NLP) and computer vision (CV) models to the scholarly domain, scientific documents remain challenging to work with. They’re often in difficult-to-use PDF formats, and the ecosystem of ...

How do people cite this paper?

(generated 20 days ago)

PaperMage has been widely adopted as a practical toolkit for parsing scientific PDFs — used to extract full texts, figures, tables, and structured content in research pipelines spanning evidence retrieval, scientific claim verification, and data extraction from research literature — while also serving as a foundational library on which new document understanding systems build their processing pipelines by leveraging its Entity/Layer abstractions and modular Recipe architecture, and is frequently situated alongside tools like GROBID as a representative pipeline-based system tailored to scientific documents in the growing landscape of document structure extraction, with some works noting its limitations in extracting complex elements like figures and tables in certain domains to motivate further research, and others drawing on its design concepts to inspire new document parsing frameworks.

Mentions

Loading PDF reader...

PaperMage: A Unified Toolkit for Processing, Representing, and Manipulating Visually-Rich Scientific Documents

Kyle Lo·Shannon Zejiang Shen·Benjamin Newman·Joseph Chee Chang
EMNLP·2023·24 citations🏆 Best Demo

AbstractDespite growing interest in applying natural language processing (NLP) and computer vision (CV) models to the scholarly domain, scientific documents remain challenging to work with. They’re often in difficult-to-use PDF formats, and the ecosystem of ...

How do people cite this paper?

(generated 20 days ago)

PaperMage has been widely adopted as a practical toolkit for parsing scientific PDFs — used to extract full texts, figures, tables, and structured content in research pipelines spanning evidence retrieval, scientific claim verification, and data extraction from research literature — while also serving as a foundational library on which new document understanding systems build their processing pipelines by leveraging its Entity/Layer abstractions and modular Recipe architecture, and is frequently situated alongside tools like GROBID as a representative pipeline-based system tailored to scientific documents in the growing landscape of document structure extraction, with some works noting its limitations in extracting complex elements like figures and tables in certain domains to motivate further research, and others drawing on its design concepts to inspire new document parsing frameworks.

Mentions

Paper

Loading PDF reader...