Countywide Address Management System Locator Redevelopment

Abstract

The County of Los Angeles is trying to modernize its Countywide Address Management System (CAMS). This project has provided an opportunity to improve an important and commonly used tool that is free to the public. In the past, the anomalous street taxonomy has presented a challenge for address matching that was identified for modifications to the CAMS locator. Investigations into these findings helped guide this project's efforts by analyzing the match rates. This project chose to pursue different match rates between various address samples and match scores as an exploration into the CAMS locator 's existing limitations. The results determined that lower match scores led to higher match rates.

Introduction

The Countywide Address Management System for the County of Los Angeles is used to manage and standardize addresses for its unincorporated areas and incorporated cities. The Los Angeles county has 88 cities. These 88 cities have 120 - 125 are unincorporated areas. It also has 4 million primary and secondary addresses. The main area that the County focuses is how to improve its CAMS locator which is an address locator that uses reference dataset to identify the best match addresses during the geocoding process in an effort to redevelop CAMS. This project explored how to improve match rates with a composite locator that is created to test against the CAMS composite locator. The address records are assigned to geospatial locations by geocoding via spatial geometry and coordinate data.

Why does Addressing matter?

"...maintaining CAMS as the single authoritative source for address information supports the critical services of County Department, such as Fire or Sheriff’s Department for CAD operations, road and other maintenance, home visits for Probation or Children and Family Services, or agricultural inspections. CAMS also supports the County indirectly by ensuring an accurate and complete Census count of all residential addresses." (County of Los Angeles Enterprise GIS, 2020).

Methods

Entirely performed in ArcGIS Pro, three methods were used for this project: preparation of address samples, creation of a test composite locator and batch geocoding. Two address samples were created: one from select zip codes (with anomalous street taxonomy identified in the Lancaster and Palmdale communities) and another that was random (comprehensively drawn from across all County of Los Angeles zip codes). As part of the test composite locator build, a "Street Address" locator was created to allow for matches outside of a street address range, which the CAMS locator's build was not structured to capture.

Data & Data Sources

• Parcel Data Address Records – County of Los Angeles Assessor Parcel Data (2019)
• CAMS Reference Files: Address Points (Points) and Street Segments (Streets); CAMS Locator Package – CAMS Data (2020)
• CAMS Survey – Survey 123 via County of Los Angeles Enterprise GIS (2021)

Results

The map represents the geocoded address samples (based on one match score and composite locator example) across the County of Los Angeles. The results illustrate that the Test composite locator had higher match rates than the CAMS composite locator.

Discussion

As expected, match rates decreased as match scores increased. For the 61,965 records for each address sample, percentage change in match rates between different match scores illustrated a uniform trend between the two address samples (except with the CAMS locator /“Zip Codes” address sample). Test composite locator matches (that the CAMS locator did not make) approximated where the address number might exist on the road segment in question. Considering there are bound to be gaps when updating CAMS datasets, this discovery might be worth exploring when reexamining the composition of the CAMS locator’s elements. Additionally, a match that is relatively close to the actual location may be better than no match at all when conducting analysis. 

Conclusion

The most significant finding of this project was that the Test composite locator produced higher match rates than the CAMS locator. Specifically, a "Street Address" locator proved instrumental to higher match rates. These discoveries aim to supplement the County's approach to redeveloping the CAMS locator in the future.