Taxonomy Explorer

The taxonomy explorer provides three complementary views of the fraud classification taxonomy: Sankey diagram, heatmap, and trend analysis.

Sankey Diagram

The default view shows a flow diagram from categories to subcategories. Each category node is connected to its subcategories by proportional flow links.

• Categories appear on the left; subcategories on the right.

• Link width indicates case volume flowing from category to subcategory.

• Use the period selector (30d, 90d, quarter, year) to change the time range.

Heatmap Exploration

Switch to the Heatmap view to see a category × time grid:

• Rows represent fraud categories.

• Columns represent time periods (configurable granularity: day, week, month).

• Cell intensity shows case count — darker cells indicate higher volume.

• Hover over a cell to see the exact count.

Use the heatmap to identify seasonal patterns or emerging fraud types.

Trend Analysis

The Trend view shows time-series sparklines per category:

• Each category gets its own colored sparkline showing case volume over time.

• Totals are displayed next to each category name.

• X-axis labels show the first and last period.

Use trends to track whether a specific fraud type is growing or declining.

Tips

• Start with the Sankey view for a structural overview, then drill into trends for temporal patterns.

• The heatmap is useful for identifying which categories spike in specific periods (e.g., holiday-related fraud patterns).

• All three views share the same period filter for consistent comparison.