Replicable design
One system copied per LGU, tailored to local needs and values
Public interface
Shows aggregated capability gaps, priority areas, planned interventions, budget summaries, and monitoring status without exposing confidential household records.
Decision-maker workspace
Mayor's office, planning, budget, barangay leaders, CSOs, evaluators, and Tech4CaFE admins can manage data, scoring, interventions, and reports.
Sheet-aligned repository
Designed around LGU profile, needs assessment, capability scores, recommended interventions, stakeholders, implementation status, and monitoring indicators.
Access mode
Public transparency view
Confidential household details are hidden. Aggregated gaps, interventions, and report summaries remain visible.
AHP from deprivation arrangements
Weights derived from how households rank their deprivations
AHP method used in this demo
Each synthetic household ranks its top deprivation arrangement. The system converts pairwise ranking preferences into an AHP-style pairwise comparison matrix, uses the geometric mean method to estimate weights, and applies those weights to capability gap and intervention prioritization.
Synthetic citizen dataset
Random big dataset for pilot demonstration
Repository status
Local browser repository
The app can run from browser storage, or load the included random big dataset from the local data repository.
0 records
Public-safe aggregation enabled
CSV and brief exports ready
City-held capability data
This demo assumes the LGU already has local data by capability, geography, household profile, stakeholder feedback, intervention needs, and monitoring indicators.
Confidential repository controls
Visible to decision-makers and admins after login. Public users only see aggregated summaries.
LGU profile
Needs assessment
Capability scores
Household capability profiles
Recommended social policy interventions
Stakeholders and implementation status
Monitoring indicators
| Household | Area | Beneficiary group | Top deprivation arrangement | Capability gap | Recommended intervention |
|---|
Explainable AI with LIME
Explain every household recommendation without exposing confidential data
Recommendation explanation
Capability recommendation
Top contributing deprivation factors
Why this intervention?
Alternative if conditions change
Prototype note: this frontend simulates LIME by perturbing synthetic household capability features and estimating local contribution weights. In production, run LIME on a backend model using validated LGU data, audit logs, and access controls.
Social policy interventions
Prioritize communities and interventions for each capability
| Intervention | Capability | Priority communities | Budget | Impact | Equity | Feasibility | Priority |
|---|
Outputs
Reports and export-ready summaries
Capability gap report
Household capability profile
Public version shows anonymized and aggregated profiles. Decision-maker version can use household-level records under access controls.