Parse and extract text from scholarly documents
Query PDF documents using natural language
Extract text from images using OCR
Search information in uploaded PDFs
A demo app which retrives information from multiple PDF docu
Convert images with text to searchable documents
Search documents for specific information using keywords
Search... using text for relevant documents
Extract handwritten text from images
Find relevant legal documents for your query
Extract and query terms from documents
Process text to extract entities and details
Traditional OCR 1.0 on PDF/image files returning text/PDF
Grobid End to End Evaluation is a tool designed to assess and validate the performance of Grobid, a machine learning model used for parsing and extracting text from scholarly documents. It evaluates the accuracy and reliability of Grobid's output by comparing it against ground truth data, ensuring the extracted text meets high standards of quality and usability.
What file formats does Grobid End to End Evaluation support?
Grobid End to End Evaluation supports PDF, XML, and plain text formats for both input documents and ground truth files.
Can I customize the evaluation metrics?
Yes, Grobid End to End Evaluation allows users to define custom metrics and weighting to suit specific requirements.
How do I handle documents with complex layouts?
Grobid is specifically designed to handle complex layouts, including multi-column text, tables, and figures. Ensure your ground truth accurately reflects these elements for proper evaluation.