About this document
Sub Word Embeddings For Address Parsing by prudhvi.konapala is a document available to read on EtoBox.
This document presents a method for multinational address parsing using a Recurrent Neural Network (RNN) architecture that leverages subword embeddings. The proposed model achieves high accuracy (around 99%) for countries used in training and demonstrates effective zero-shot transfer learning capabilities for parsing addresses from other countries. An open-source implementation of the trained models is also provided, contributing to advancements in address parsing technology.
- Author
- prudhvi.konapala
- Language
- EN