The G.A.R.D.E.N. philosophy centers on democratizing access to complex data systems through thoughtful design and automation. Rather than requiring users to master specialized query languages or understand database internals, we create intuitive interfaces that align with natural cognitive patterns. This approach makes powerful data exploration accessible to everyone, regardless of technical background.
G.A.R.D.E.N. is an open-source initiative focused on creating accessible, Python-based graph data applications for everyone. This project leverages the Module Generator to rapidly develop intuitive interfaces to graph databases, transforming complex network data into approachable, usable tools.
We believe powerful data exploration should be available to everyone, not just those with specialized technical knowledge. The G.A.R.D.E.N. philosophy embraces these core principles:
- Cognitive Alignment: Design interfaces that match how people naturally think about and explore information
- Technical Abstraction: Hide complex implementation details behind intuitive interfaces
- Rapid Development: Use automation to accelerate building and deploying data exploration tools
- Progressive Discovery: Enable users to start simply and gradually access more sophisticated capabilities
- Multiple Perspectives: Support different cognitive approaches to exploring the same data
Our approach centers on making graph data accessible without requiring extensive knowledge of graph databases or query languages. Each application in the G.A.R.D.E.N. ecosystem is designed to expose different interaction patterns with graph data while maintaining simplicity and usability.
Grassroots implements a "metadata-first" approach to graph exploration. It begins with the schema and works its way to the data, making it ideal for users who understand their business domain but may not know specific data points.
This Flask API exposes schema information through simple endpoints. By following schema information, users discover data that matches specific patterns. The interface features a clean, hyperlink-based UI without complex visualizations.
The Module Generator accelerates development by creating intuitive connections between URLs (containing labels) and the underlying Python functions. Users explore entities by examining schema information, sorting by properties, and browsing ordered lists – offering an alternative to property-based search.
Grasshopper takes a "data-first" approach to graph exploration. Users begin with a curated list of high-value entities and navigate through the graph by "hopping" between connected nodes.
Each entity has a simplified profile page showing all properties and connected nodes. Navigation happens through hyperlinks rather than visualizations. This approach creates an intuitive browsing experience similar to exploring Wikipedia, where each click reveals new connections.
Grasshopper excels at providing contextual understanding of relationships and allowing serendipitous discovery through graph traversal.
Sunflower offers a "pattern-first" approach to graph exploration, focusing on revealing common relationship patterns within your data. This application automatically identifies and categorizes recurring structural patterns (like cycles, stars, or chains) that may have business significance.
Users begin by selecting a pattern type from a categorized list. Sunflower then displays all instances of that pattern in the database, allowing users to browse examples and understand how entities commonly relate to one another. Each pattern instance can be explored in detail through hyperlinked entity profiles.
This approach excels at revealing hidden structural patterns that might indicate important business rules, constraints, or opportunities – all without requiring complex queries or visualizations.
To begin exploring the G.A.R.D.E.N. ecosystem, start with our Prerequisites notebook. This introduction covers the fundamental concepts needed to understand and extend the applications in this repository.
All G.A.R.D.E.N. applications are built using the Module Generator, which automatically creates type-safe Python interfaces for Neo4j databases. This approach offers several advantages:
-
Rapid Development: Applications can be developed in days rather than weeks by leveraging auto-generated database interfaces.
-
Consistent Patterns: All applications share consistent interaction patterns, reducing the learning curve for users and developers.
-
Extensibility: The modular architecture makes it easy to enhance applications with new features while maintaining a solid foundation.
G.A.R.D.E.N. applications can be adapted for numerous domains:
-
Government Data Transparency: Create public-facing portals for exploring relationships between policies, programs, and outcomes.
-
Research Data Exploration: Help researchers discover unexpected connections between entities in scientific datasets.
-
Enterprise Knowledge Management: Build internal tools for navigating organizational knowledge and relationships.
-
Educational Tools: Create interactive environments for students to explore complex domains through their relationship structures.
We welcome contributions of all kinds, from bug fixes to entirely new "garden" applications. If you're interested in contributing, please review our contribution guidelines and code of conduct.
The G.A.R.D.E.N. project aims to make graph data accessible to everyone through simple, intuitive interfaces. Join us in growing this ecosystem of tools for graph exploration and discovery!
