Header Extraction from Scientific Documents

With the massive amount of published material becoming accessible to the public via the World Wide Web, a tool that can parse header information from research papers will be invaluable to systems concerned with storing and retrieving scientific publications. Given certain search criteria and metadata, we would like to have some way of finding and identifying a document that matches our aforementioned criteria. A simple way of identifying an article is by header information (Fig 1), e.g. title, authors, affiliation, published date, etc. Thus, our goal is to put together a tool that can efficiently and accurately extract header data from research papers.