Talk Proposal Submission
If you are interested in attending this talk at PyCon JP 2017, please use the social media share buttons below. We will consider the popularity of the proposals when making our selection.
poster
AddressExtract: Automatically extracting postal addresses from the Web(en)
Speakers
Michael Penkov
Audience level:
Intermediate
Category:
Industry Uses
Description
AddressExtract is a closed-source library that I have been working on for several years. The goal of this session is to report on results, obtain feedback, and gauge interest for potentially open-sourcing the library in the future.
Objectives
Abstract
AddressExtract (AE) is a Python library for automatically extracting postal addresses from the Web. It achieves this by leveraging conventional machine learning techniques (SVM).
This presentation describes how AE works and reports on its results. I will explore the challenges of extracting addresses from around the world, and present my current solutions.
