BusinessCardParser

This program is a business card reader with the following attributes

• Uses parsed results from Optical Character Recognition (OCR)
• Extracts name, phone number and email address from business card
• Provides sample test cases and sample main() output using the Input.txt file

How to Setup

• Install Natural Language Toolkit (NLTK) from Install NLTK Module
• Run most recent version of Python 3.*

Programming Approach

• Names were extracted from the document using the nltk named entity chunker
  • This chunker assigns values of PERSON, ORGANIZATION, TIME, etc. to each chunk of information
  • A reference used for NLTK is NLTK Reference
  • NLTK assigns job titles as a chunk of PERSON
    • A running list of common job keywords was created and is checked before accepting the information as chunked as PERSON as a name
    • This list should be added to based on positions available, but a few examples were given in this code
• Phone numbers were extracted using regular expressions
  • A reference used for regular expressions for phone number is Phone Number RegEx
  • The regular expression was then customized for this application
  • A good tool for solving regular expressions is https://www.debuggex.com/
• Email addresses were also extracted using regular expressions
  • A reference used for regular expressions for email addresses is Email Address RegEx
  • This regular expression was then customized for this application

Interface Specification

• The method get_contact_info(document) will return an instance of the object ContactInfo(name, email, phone)
• The attributes can then be attained using the respective get methods in the ContactInfo class