Alabama NENA Conference Reflection – Street Name Parsing and Normalization

November 6, 2024

My previous article reflected on the process of synchronizing MSAG and Road Centerline data which of course can be painfully time-intensive and tedious. This article focuses on another challenging task: parsing street names, an even broader and more complex issue.

Converting legacy street names to a modern format can be more complex than it appears. If parsing the four-part legacy format and standardizing it to the eight-part NENA format isn’t challenging enough, factors like cultural variations in street names (e.g., Spanish, French, German), unconventional prefixes, and inconsistent or poor naming conventions can make anyone’s head spin!

This requirement brought back nightmares when I was in the trenches battling against legacy to NG street name parsing in Yuma, Arizona, where Spanish language influence was significant.

To address this, we developed AI-based parsing and standardization functions, code named A3i. After seeing how the ai3 function performed in GIVE during the conference several Alabama practitioners have commented, “That would have saved us a year!”

Indeed, our exclusive A3i technology was designed to eliminate this headache.A3i parses and standardizes translation between E9-1-1 and NG9-1-1 (and vice versa). The GIVE UI suggests changes for both four- and eight-field standards for review and approval, all with the help of Machine Learning behind the scenes!

When a street name that A3i does not fully recognize it’s automatically flagged, logged, and monitored by our engineers. These flagged instances are used to create additional training datasets, enhancing A3i’s ability to recognize new patterns in the future. GIVE guarantees a 99% accuracy rate in parsing to all of its subscribers.