NSF Proto-OKNBeginnerEnded
NSF Gaps-to-Graphs (DREAM)
“⭐ Enhance and Accelerate City & Shelter Homeless Capabilities with AI Knowledge”
The Challenge
What you'll do: • Research and analyze the problem space • Explore gaps in AI-powered data models • Present your approach with a proposal for sources and how to fix the gap What you'll deliver: A list of gaps + options to remedy. Why it matters: Your Challenge: Enhance +20% and Accelerate x10 AI Research’s Impact on Software Supply Chains For an hour, attendants will help Proto-OKN projects transition to markets, by co-piloting data enhancement solutions for their Knowledge Graphs. Developing AI requires both technical and domain/business leaders to collaborate. NO TECHNICAL SKILLS NEEDED
Timeline
New Event
Mentor SessionFri, Mar 27, 2026
This is when we will gather, test, and co-create.
About This Challenge
Discover and address critical gaps in homeless services knowledge to strengthen support systems, improve outcomes, and enhance coordinated response across cities. This master-level challenge focuses specifically on the DREAM Proto-OKN Knowledge Graph - an advanced AI system mapping homelessness services, interventions, and outcomes into an interconnected knowledge network. You'll work directly with urban researchers and homeless service providers to identify where this knowledge system falls short and develop strategies to fill those gaps. ## What You'll Explore The DREAM Knowledge Graph captures critical intelligence about: - Emergency shelter capacity, utilization patterns, and service delivery models - Housing placement success rates and barriers to permanent housing transitions - Mental health and substance abuse intervention effectiveness - First responder protocols and resource coordination during homeless-related emergencies - Case management workflows and client outcome tracking across service providers - Weather-related emergency response and vulnerable population protection strategies Your Mission: Identify where this research-grade knowledge graph has gaps in knowledge or impact, and develop practical options to remedy these deficiencies affecting both homeless individuals and the first responders who serve them. ## The Homeless Services Knowledge Challenge Current homeless services knowledge systems face significant limitations: - Fragmented Data: Critical client information and service histories remain isolated across shelters, healthcare, and social services - Response Delays: First responders and case workers lack real-time access to available beds, services, and client histories - Outcome Blind Spots: Disconnected systems miss successful intervention patterns and service gaps that perpetuate homelessness cycles - Coordination Failures: Knowledge systems struggle to support the complex multi-agency responses required for effective homeless services The DREAM Proto-OKN aims to address these issues, but needs expert analysis to identify blind spots and enhancement opportunities. ## Your Analytical Role No technical experience required - we need your homeless services domain expertise and frontline insight. You'll work in focused teams including: - DREAM Proto-OKN researchers and data scientists - Homeless shelter directors and case management staff - Police officers, paramedics, and emergency response coordinators - Social workers and mental health professionals - City housing officials and homeless services administrators - Formerly homeless individuals and community advocates Together, you'll: 1. Identify Knowledge Gaps: Pinpoint what critical homeless services information is missing or inadequately represented 2. Assess Coordination Failures: Determine where the knowledge graph fails to support effective multi-agency response 3. Develop Data Solutions: Design practical approaches to capture and share essential client and service information 4. Prioritize Integration Opportunities: Rank gap-filling strategies by impact on homeless outcomes and first responder effectiveness 5. Map Implementation Pathways: Connect gap analysis to actionable improvement roadmaps for cities and service providers ## Real Impact Potential Your contributions will directly influence: - Enhanced Emergency Response: Improved data access enabling faster, more informed responses to homeless-related emergencies - Better Service Coordination: Connected systems reducing duplicate services and missed intervention opportunities - Stronger Housing Outcomes: More complete knowledge improving pathways from street to stable housing - First Responder Effectiveness: Better situational awareness and resource coordination for police, EMS, and outreach teams ## Success Metrics Core Deliverables: - Comprehensive gap inventory identifying knowledge and coordination deficiencies - Prioritized remediation options with implementation feasibility assessments for cities and service providers - Impact analysis quantifying how gap-filling improves homeless outcomes and emergency response - Actionable roadmap for DREAM Proto-OKN enhancement tailored to urban homeless services Broader Impact: - Shelter staff gain real-time visibility into citywide capacity and client needs - First responders access critical background information improving safety and service delivery - Case managers coordinate more effectively across agencies reducing service gaps - Cities make data-driven decisions about homeless services resource allocation and policy ## Getting Started Come prepared to share your expertise about homeless services challenges, coordination gaps, and data limitations in your domain. Whether you work in emergency shelters, street outreach, police response, emergency medical services, housing navigation, or city administration - your perspective is essential for identifying where our knowledge systems fall short. Homeless services knowledge gaps perpetuate cycles of homelessness and strain emergency response systems. Join us to find them and fix them.
Research Team
Principal Investigator: Dr. Yuzhou Chen and collaborative research team developing the DREAM (Data-driven Real-time Emergency response And Management) Knowledge Graph under the NSF Proto-OKN initiative. ## The DREAM Proto-OKN Research Initiative This challenge connects you directly with cutting-edge research funded through NSF's Prototype Open Knowledge Network program - a multi-million dollar initiative to create interconnected knowledge systems that address critical societal challenges. The DREAM project specifically targets the fragmented landscape of homeless services through advanced knowledge graph technologies. ### Research Team Expertise Lead Research Group: Dr. Yuzhou Chen's team brings together specialists in: - Knowledge Graph Engineering: Advanced semantic web technologies and ontology development for social services - Real-time Systems: Emergency response coordination and dynamic resource allocation - Urban Informatics: Data integration across fragmented municipal and nonprofit service systems - Human-Centered AI: Ensuring knowledge systems serve both service providers and vulnerable populations ### Multi-Institutional Collaboration The DREAM Proto-OKN represents a convergence research approach, integrating: Technical Partners: - Computer science researchers developing scalable knowledge representation - Data scientists specializing in heterogeneous social services datasets - HCI experts focused on crisis response interfaces Domain Partners: - Homeless service providers and case management organizations - Emergency management agencies and first responder departments - Municipal housing authorities and coordinated entry systems - Community health centers and mental health providers ### Research Innovation Focus The team is pioneering next-generation knowledge graphs that can: - Dynamically integrate real-time shelter capacity, weather alerts, and emergency response data - Learn from intervention patterns to identify successful pathways out of homelessness - Coordinate multi-agency responses during weather emergencies and public health crises - Preserve privacy while enabling critical information sharing across service boundaries ### Your Role in Active Research As a challenge participant, you're not just analyzing existing data - you're contributing to ongoing NSF-funded research that will shape how cities nationwide respond to homelessness. Your gap analysis and remediation strategies will directly inform the next phase of DREAM system development and deployment. This is applied research with immediate real-world impact, where your insights help strengthen the knowledge infrastructure supporting society's most vulnerable populations.
Resources
# Data Sources for NSF Gaps-to-Graphs (DREAM) Challenge ## Federal Government Datasets • HUD Point-in-Time (PIT) Count Data - Annual homeless population counts by CoC - https://www.hudexchange.info/resource/3031/pit-and-hic-data-since-2007/ - Geographic distribution, demographics, sheltered vs unsheltered populations • HUD Housing Inventory Count (HIC) - Comprehensive shelter and housing capacity data - https://www.hudexchange.info/resource/4832/2022-hic-data-by-coc/ - Bed capacity, utilization rates, program types by Continuum of Care • HUD Longitudinal Systems Analysis (LSA) - Client-level outcomes and service patterns - https://www.hudexchange.info/programs/lsa/ - Length of stay, housing placement rates, service utilization • SAMHSA Treatment Episode Data Set (TEDS) - Substance abuse treatment admissions - https://www.datafiles.samhsa.gov/dataset/treatment-episode-data-set-admissions-2020-teds-2020-ds0001 - Mental health and addiction services for homeless populations • CDC WONDER - Mortality data and health outcomes by geographic area - https://wonder.cdc.gov/ - Death rates, causes of death for vulnerable populations ## State and Local Government Sources • Homeless Management Information System (HMIS) - Local CoC client tracking systems - Contact individual Continuums of Care for data sharing agreements - Real-time bed availability, client histories, service coordination • Emergency Medical Services (EMS) Call Data - 911 response patterns - City fire departments and emergency services departments - Frequent utilizer patterns, response locations, medical emergencies • Police Computer Aided Dispatch (CAD) Data - Law enforcement homeless encounters - Local police departments' open data portals - Incident locations, response types, resource allocation • 311 Service Request Data - Citizen reports of homeless encampments and services - Most major cities publish via open data portals - Geographic clustering, service gaps, community concerns ## Research Repositories and Academic Sources • National Alliance to End Homelessness Research Library - https://endhomelessness.org/resource/ - Best practices, intervention effectiveness studies, policy research • Corporation for Supportive Housing Knowledge Hub - https://www.csh.org/knowledge-hub/ - Housing First outcomes, supportive services research • Urban Institute Data Catalog - Housing and homelessness research datasets - https://datacatalog.urban.org/search/type/dataset - Cost studies, program evaluations, demographic analyses • Inter-university Consortium for Political and Social Research (ICPSR) - https://www.icpsr.umich.edu/web/pages/ - Longitudinal studies on homelessness, housing instability • National Low Income Housing Coalition Research Database - https://nlihc.org/research - Affordable housing availability, rent burden data by metro area ## Healthcare and Mental Health APIs/Databases • Healthcare Cost and Utilization Project (HCUP) - https://www.ahrq.gov/data/hcup/index.html - Emergency department utilization, frequent users, mental health diagnoses • National Institute of Mental Health Data Archive - https://nda.nih.gov/ - Mental health intervention outcomes, treatment effectiveness • Behavioral Risk Factor Surveillance System (BRFSS) - https://www.cdc.gov/brfss/data_documentation/index.htm - Population-level mental health and substance use indicators ## Weather and Environmental Data • National Weather Service API - Real-time and historical weather data - https://www.weather.gov/documentation/services-web-api - Extreme weather events affecting vulnerable populations • EPA Air Quality System (AQS) API - Environmental health factors - https://aqs.epa.gov/aqsweb/documents/data_api.html - Air quality impacts on outdoor homeless populations ## Economic and Social Context APIs • Census American Community Survey (ACS) API - Demographics and housing costs - https://www.census.gov/data/developers/data-sets/acs-1year.html - Rent burden, poverty rates, population characteristics • Bureau of Labor Statistics API - Employment and economic indicators - https://www.bls.gov/developers/ - Job availability, wage trends, economic conditions by metro area • Opportunity Insights Economic Mobility Data - https://opportunityinsights.org/data/ - Neighborhood-level economic opportunity indicators ## Specialized Homelessness Research Sources • National Center on Homelessness Among Veterans (NCHAV) - https://www.va.gov/homeless/nchav/ - Veteran-specific homelessness data and intervention outcomes • National Center for Homeless Education Data Portal - https://nche.ed.gov/data-and-stats/ - Homeless student demographics, educational outcomes • Homebase Research Database - Prevention and rapid rehousing studies - https://homebaseccc.org/what-we-do/research-evaluation/ - Cost-effectiveness analyses, program evaluation data ## Real-Time Service Coordination Platforms • Coordinated Entry System Data - Prioritization and housing placement tracking - Local CoCs maintain systems like CES or VI-SPDAT databases - Assessment scores, housing placement wait times • 211 Service Directory APIs - Social services inventory and referrals - https://www.211.org/about/api - Available services, capacity, eligibility requirements • United Way Resource Directory APIs - Community service mappings - Local United Way chapters often provide API access - Service provider networks, program descriptions ## Longitudinal Research Studies • Panel Study of Income Dynamics (PSID) - Long-term housing instability tracking - https://psidonline.isr.umich.edu/ - Multi-generational housing security patterns • Survey of Income and Program Participation (SIPP)
Skills
You'll Learn
Data AnalysisStrategic PlanningPresentation SkillsSoftware DevelopmentRequirements GatheringStakeholder ManagementDisruptive InnovationBusiness AnalysisOpen Innovation
Helpful to Have
Outcomes Research
Hosted by
Hugo
0 teams / 50 max
Starts Mar 27, 2026, 12:30 PM (EDT)
Ends Mar 27, 2026, 1:30 PM (EDT)
Leaderboard
On this page
Prizes
- Certificates, Badges.