Farmers-market document extraction
A pipeline that turns HTML, PDF, Markdown, and scanned documents about farmers markets into a structured database.
- Role
- Undergraduate Research Assistant
- Stack
- Python, HTML parsing, PDF parsing, Multimodal OCR
An accessibility index for city planning needs clean data on each farmers market. That data is in many web pages, PDF files, and scanned flyers.
This pipeline parses HTML, PDF, and Markdown sources about farmers markets. It writes each market into a structured database.
Some market documents are scans or images, so they have no text layer. Multimodal OCR tools read the text from these documents and turn it into structured records.