Features of Global Address Parser

The module has these features:

• Splits and formats address strings into its components with the help of models trained through machine learning.
• Parses addresses in Roman script and accepts input addresses in Roman script. Few Greek alphabets are also supported.
• Currently supports parsing for three countries, Germany, United Kingdom, and Canada.
• Handles country-specific addressing conventions efficiently. Address components of different countries vary in many ways. For example, in German addresses house number usually comes after the street name and the post code comes before the city. The module handles all these complexities efficiently and predicts the address components in accordance with the conventions of the specified country.
• Does away with the need of reference address databases for parsing.