Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Understanding Housing and Homelessness System Access by Linking Administrative Data

Geoffrey G. Messier, Sam Elliott, Dallas Seitz

TL;DR

This work tackles linking fragmented housing and homelessness records under privacy constraints to reveal system-wide interactions without exposing identifiable information. It proposes a two-step approach: privacy-preserving pairwise profile matching using Bloom-filter-encoded identifiers and the Dice coefficient, followed by clustering to form latent individuals, evaluated with both traditional linkage metrics and domain-specific HHSC utilization metrics. Key findings show that privacy-preserving linkage is practical for HHSC data, with domain-specific metrics amplifying observed differences between linkage methods; a simple threshold-based approach coupled with efficient clustering often rivals more complex models in real-world utility. The study demonstrates that linked data significantly alters interpretations of service use (e.g., shelter stays and tenure) and provides actionable insights for HHSC operators while preserving resident privacy. This has practical implications for data integration in fragmented public service systems and informs policy design without compromising privacy.

Abstract

This paper uses privacy preserving methods to link over 235,000 records in the housing and homelessness system of care (HHSC) of a major North American city. Several machine learning pairwise linkage and two clustering algorithms are evaluated for merging the profiles for latent individuals in the data. Importantly, these methods are evaluated using both traditional machine learning metrics and HHSC system use metrics generated using the linked data. The results demonstrate that privacy preserving linkage methods are an effective and practical method for understanding how a single person interacts with multiple agencies across an HHSC. They also show that performance differences between linkage techniques are amplified when evaluated using HHSC domain specific metrics like number of emergency homeless shelter stays, length of time interacting with an HHSC and number of emergency shelters visited per person.

Understanding Housing and Homelessness System Access by Linking Administrative Data

TL;DR

This work tackles linking fragmented housing and homelessness records under privacy constraints to reveal system-wide interactions without exposing identifiable information. It proposes a two-step approach: privacy-preserving pairwise profile matching using Bloom-filter-encoded identifiers and the Dice coefficient, followed by clustering to form latent individuals, evaluated with both traditional linkage metrics and domain-specific HHSC utilization metrics. Key findings show that privacy-preserving linkage is practical for HHSC data, with domain-specific metrics amplifying observed differences between linkage methods; a simple threshold-based approach coupled with efficient clustering often rivals more complex models in real-world utility. The study demonstrates that linked data significantly alters interpretations of service use (e.g., shelter stays and tenure) and provides actionable insights for HHSC operators while preserving resident privacy. This has practical implications for data integration in fragmented public service systems and informs policy design without compromising privacy.

Abstract

This paper uses privacy preserving methods to link over 235,000 records in the housing and homelessness system of care (HHSC) of a major North American city. Several machine learning pairwise linkage and two clustering algorithms are evaluated for merging the profiles for latent individuals in the data. Importantly, these methods are evaluated using both traditional machine learning metrics and HHSC system use metrics generated using the linked data. The results demonstrate that privacy preserving linkage methods are an effective and practical method for understanding how a single person interacts with multiple agencies across an HHSC. They also show that performance differences between linkage techniques are amplified when evaluated using HHSC domain specific metrics like number of emergency homeless shelter stays, length of time interacting with an HHSC and number of emergency shelters visited per person.
Paper Structure (20 sections, 1 equation, 1 figure, 11 tables)

This paper contains 20 sections, 1 equation, 1 figure, 11 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Data Set
  4. Synthetic Data Generation
  5. Linkage and Clustering Methodology
  6. Dice Coefficient Calculation
  7. Pairwise Match Training and Test
  8. Pairwise Match Models
  9. Threshold
  10. Logistic Regression (LR)
  11. Decision Tree
  12. Multi-layer Perceptron (MLP)
  13. Clustering Algorithms
  14. Housing/Homelessness System of Care Utilization Analysis
  15. Results
  16. ...and 5 more sections

Figures (1)

  • Figure 1: Shelter access episode length distribution.