Visualise Data Relationships

How to use

1. Choose JSON, YAML or CSV, then paste, type or load your data from a local file or the built-in sample.
2. Map a Source field or column and a Target field or column, and an optional Label.
3. Check the relationship graph, or the relationship table and text view alongside it.
4. Download the graph as SVG or PNG if you want to keep or share it.

Worked example

Load the built-in sample to see a small dependency graph — an app calling an API, the API reading and writing a database, and a worker writing to the same database — with the Source, Target and Label fields already mapped, since the sample uses those exact field names.

About the views

Graph, table and text

The graph is a deterministic diagram built only from the relationships you mapped, laid out the same way every time for the same data. The relationship table and the text view show exactly the same relationships as plain rows and lines, so nothing in the graph depends on being able to see it.

Directional relationships

Every relationship points from its source to its target, shown as an arrow in the graph and as separate Source and Target columns in the table. A relationship pointing back the other way, or a record whose source and target are the same, is shown as its own distinct relationship.

Duplicates, cycles and self-loops

A repeated relationship is never merged or counted — each one you mapped is drawn and listed on its own. Relationships that form a loop back to an earlier record, or that point from a record to itself, are fully supported.

Limitations

• Relationships are never guessed. Only the exact field or column you map as Source and Target is used; a field named similarly, such as "from" or "parent", is never mapped automatically.
• One record with an invalid or missing Source or Target stops the graph, table and text view from being built at all, rather than leaving that record out silently. Fix or remove the record to see the rest.
• JSON and YAML sources must be a top-level list of records; a nested field path, such as a value inside another object, cannot be mapped in this version.
• The graph's node labels may be shortened to fit; the table and text view always show the complete text.
• Characters that cannot appear directly in an SVG file are shown using visible Unicode escape notation, such as \u0000, in the graph and in downloaded images. The relationship table and text view always show the original text.
• A very large graph may become too complex to draw as an image; the relationship table and text view remain available either way.
• There is no dragging, rewiring or editing. The graph is a read-only view for inspecting and exporting relationships you have already mapped.

Privacy and accessibility

Privacy

Your source and generated views stay on this device during normal use. MeaningKit does not intentionally send source content, filenames or image exports to its servers for processing. Downloading an image is your choice. Read the privacy page.

Accessibility

The graph is hidden from assistive technology, since the relationship table and text view already give the complete relationship information in ordinary, readable form. Every control, including the field mapping and the SVG and PNG downloads, works with a keyboard. Read the accessibility page.