README.md

In dsidavis/CVRead: Extract information from Curriculum Vitae/Resume in PDF

This is a package that explores reading PDF files for curricula vitae or resume and to extract information from these.

The file Anatomy provides some thoughts and approaches to identifying the different pieces.

We are currently using the Python module pdfminer to read the PDF and emit information about its content in XML. This contains the font and bounding box information. From these, we can recover a lot of the information and guess the nature of different text using XPath queries and manipulation in R.

The package uses new features (October 2015) in the XML package. You can install it (from source) via

    install.packages("XML", repos = "http://www.omegahat.org/R")

The initial version of the package provides a variety of different ways to get at the information we want in the PDFs. These represent different strategies for identifying that information (see the file Anatomy). It is unlikely that one of these strategies will work for all PDFs. Instead, we will have to learn which order to apply them, how to determine if they are wrong and move to another, and to add new strategies as we encounter new patterns. This extensibility is very important, as is the ability to determine whether a set of results is correct for a given PDF.

To convert a PDF document to XML in the format we want, you will need to install pdfminer http://www.unixuser.org/~euske/python/pdfminer/. The script we want is pdf2txt.py. You can place this anywhere and add the directory to your shell's PATH variable. Alternatively, you can specify its location using the R option PDF2TXT, e.g.,

options(PDF2TXT = "PythonTools/pdfminer/tools/pdf2txt.py")

dsidavis/CVRead documentation built on May 15, 2019, 4:19 p.m.