DATA.GOV.HK

Project Background

Hong Kong’s Central Open Data Catalogue

DATA.GOV.HK serves as Hong Kong’s official open data gateway, consolidating datasets from government bureaux, departments, and public bodies into a single discoverable catalogue.

With thousands of datasets, over a hundred provider organisations, API-enabled resources, and sustained reuse across sectors, API-enabled resources, and sustained public reuse, the platform operates as infrastructure rather than a conventional website. Its role is to enable structured data discovery, reliable download access, and long-term dataset governance across a growing ecosystem of contributors and users.

The revamp strengthened discoverability, refined retrieval logic, and modernised key interaction models while preserving operational continuity.

Our Approach

Cross-Stakeholder Behaviour Mapping

We synthesised discovery patterns across data providers and data users into a unified behavioural framework. The research shaped structural decisions in taxonomy, metadata hierarchies, and retrieval logic, aligning contributor governance constraints with user discovery expectations.

Information Architecture & Taxonomy Restructuring

Dataset categories, provider structures, and metadata presentation were restructured to reduce ambiguity and improve navigational predictability. The goal was structural clarity across thousands of datasets without oversimplifying filtering depth.

Search & Filtering Engineering

Search and multi-dimensional filtering were treated as core discovery infrastructure. Persistent query access, refinable result controls, and clearer result hierarchy support both targeted API retrieval and exploratory browsing across large dataset volumes.

Historical Data & Retrieval Logic

The platform’s historical dataset mechanism was refined to support structured access to archived data versions. Retrieval flows were engineered to manage long-horizon data accumulation while maintaining predictable user interaction.

Platform Modernisation & API Visibility

The revamp strengthened API discoverability, download management, and extensibility without disrupting operational stability. Engineering updates were implemented under governed change control to ensure continuity and future scalability.

Public UX Workshops

Evidence-Led Design Across Providers and Users

Four structured UX workshops were conducted with bureaux and departments, researchers, programmers, and members of the public. Each group surfaced distinct friction points — from metadata inconsistency and search refinement challenges to historical data access and download control.

Insights were consolidated into prioritised design directives, ensuring that architectural changes were grounded in validated behavioural evidence rather than assumptions.

See How It Works

Platform Capabilities

Designed for Discoverability and Reuse

Intuitive + Extensive FilteringFiltering was redesigned as core discovery infrastructure rather than a secondary interface feature. Multi-dimensional controls allow refinement by category, provider, format, and API availability, while persistent search access and visible filter states reduce disorientation. The objective was structured result control across large datasets, enabling precise narrowing without repeated query attempts.

Historical Data DownloadThe platform supports structured access to archived dataset versions alongside current releases. Historical retrieval mechanisms were refined to maintain version clarity and predictable navigation, allowing users to access prior data iterations without manual file tracing or ambiguous naming structures.

Dataset BookmarkingBookmarking enables users to save frequently referenced datasets within a structured personal dashboard. This supports recurring research and development workflows, reducing repeated discovery effort and maintaining continuity across sessions without disrupting the primary browsing experience.

Download QueueA controlled download queue was introduced to manage multi-file and high-volume retrieval. Users can batch selected resources, review queued items, and initiate organised download sequences, reducing fragmented actions and supporting more systematic data extraction workflows.

Filtering was redesigned as core discovery infrastructure rather than a secondary interface feature. Multi-dimensional controls allow refinement by category, provider, format, and API availability, while persistent search access and visible filter states reduce disorientation. The objective was structured result control across large datasets, enabling precise narrowing without repeated query attempts.