Questions? We have answers.

Expanded entry details in TRA analyses

  • Updated

Summary

The expanded entry details page in Transaction Risk Analytics (TRA) provides comparative visuals and additional context to help interpret risk scores and make informed analysis decisions

Learn about the expanded entry details page available in the results of MindBridge's Transaction Risk Analytics.


Go to the expanded entry details

  1. On the Data page, select View analysis in line with the analysis you want to see.
  2. Go to the Data table tab.
  3. Select the chevron icon ( Expand icon ) in line with the data you want to drill into.
    Any relevant entries will be exposed.
  4. Select any of the exposed entries to drill into it.
    You will go to the entry details page, which shows data from the entry you selected at the top of the page, and provides risk score context below.
Tip: Open the Columns menu (on the right side of the page) and use the checkboxes to select which columns you want to see in the table at the top of the page.

Risk scores

The risk scores on the left side of the page are determined by the configuration you are using, as they each aggregate the results of a different combination of relevant control points. For example, the Payroll Analysis template provides risk scores focused on employee-based analytics, whereas the Vendor Invoice Analysis template offers risk scores focused on vendor-based analytics.

You also have the option to work with MindBridge to configure risk scores for your use case. Reach out to your CSM for details.

Note: Overall Risk is the default risk score and will be present in most TRA analyses.

Risk score histogram

Select a risk score to update the graph at the top of the page.

Note: Overall Risk is selected by default.

Entries are sorted by risk range and grouped into Low, Medium, and High risk. Each bar represents the number of entries that occurred within the indicated risk range, and shows the distribution of the risk score across the dataset.

The selected entry is highlighted in light blue to show how it was scored relative to the rest of the population in the data. Selected entry details are always present:

  • Entry risk score: The aggregate score of all control points associated with the selected risk score, based on the selected entry.
  • Risk range: The minimum and maximum risk percentage that this group of entries falls within.
  • Entries in range: The number of entries with a risk score that fell between the indicated range.
  • Total entries: The total number of entries in the dataset.

Hover over a bar to see additional details about the entries within it:

  • Risk range: The minimum and maximum risk percentage that this group of entries falls within.
  • Entries in range: The number of entries with a risk score in the indicated range.
  • Total entries: The total number of entries in the dataset.

Screenshot_showing_the_risk_score_histogram_for_overall_risk.png

 


Control points

The risk scores mentioned above leverage different combinations of relevant control points. When you select a risk score, you can see details about the various control points that contribute to the score by scrolling down the page.

Select a control point to update the graph within this section.

Note: The highest risk control point is selected by default.

Control point graphs

Within each of these graphs, the selected entry is highlighted in light blue to show how it was scored relative to the rest of the population in the data.

Tip: You can drag the selected entry window around within the graph if it is in the way of other data points.

Depending on the control point you select, the data will be presented in one of the following graph types.

Histogram

This graph indicates how many times a range of values appears (across the dataset or within a given segment of data) by grouping entries by risk range and sorting them into Low, Medium, and High risk sections. 

Selected entry details are always present:

  • Entry control point score: The risk score associated with the selected control point.
  • Risk range: The minimum and maximum risk percentage that this group of entries falls within.
  • Entries in range: The number of entries with a risk score that fell between the indicated range.
  • Total entries: The total number of entries in the dataset.

Hover over a bar to see:

  • Amount range: The minimum and maximum amounts detected, or risk scores, in the dataset or the given segment of data.
  • Entries in range: The total number of entries that fall within the range.
  • Total entries: The total number of entries in the dataset or the given segment of data.

Generally, entries with higher risk scores will appear in less-populated areas of the graph.

Screenshot_showing_a_histogram_for_the_selected_control_point.png

Scatterplot with range

This graph plots values over time, based on all the data in the dataset or a given segment of data. 

The line represents the rolling median of average amounts in a moving time window (as determined during configuration), and the shaded range represents either the rolling standard deviation of average amounts for the given time period, or the static standard deviation of residuals across the span of time in the given segment. 

To adjust the time window displayed, click and drag the time slider to the desired range.

Use the legend below the graph to toggle the different visuals off and on.

Selected entry details are always present:

  • Time period: The period of time over which the selected entry occurred. This is indicated in the first detail row, and is determined during TRA configuration. 
  • Entry control point score: The risk score associated with the selected control point.
  • Primary identifier: The main characteristic assigned to the data, determined during TRA configuration (for example, Employee ID).
  • Entry amount: The amount associated with the selected entry. This could be a monetary value, an item count, or something else, and is determined during TRA configuration.

Hover over the line to see:

  • Rolling median: The median amount in a given time window.
  • Maximum range: The maximum expected amount for the given time period.
  • Minimum range: The minimum expected amount for the given time period.

Dots indicate groups of entries in the dataset.

Hover over a dot to see:

  • Time period: The period of time over which the entry occurred. 
  • Data table filter: This label will change depending on the view you selected on the Data table dashboard.
  • Entry amount: The amount present in the entry.

Screenshot_showing_a_scatteplot_for_the_selected_control_point.png

Line plot with range

This graph plots the activity of the given segment over time, as well as the average activity and the expected activity range.

The dark blue line represents the entry count for the given time period.

Use the legend below the graph to toggle this graph's visuals on and off.

Selected entry details are always present:

  • Time period: The period of time over which the selected entry occurred. This is indicated in the first detail row, and is determined during TRA configuration. 
  • Entry control point score: The risk score associated with the selected control point.
  • Entry count: The number of entries that occurred during the given time period.
  • Average amount: The mean of the count of activity in the given segment of data.
  • Maximum range: The top of the range represents the value of the light blue line +1 standard deviation.

Hover over a dot on the dark blue line to see:

  • Time period: The period of time over which the entry occurred. 
  • Entry count: The number of entries that occurred during the given time period.
  • Maximum range: The maximum expected amount for the given time period.
  • One of the following, based on your TRA configuration:
    • Average amount: The mean of the count of activity in the given segment of data.
    • Rolling median: The median count of activity in the given time window.
    • Nth percentile: The percentile (determined during configuration) of the count of activity in the given time window.

The light blue line represents either 1) the mean of the number of entries in the time period for the given segment of data, 2) the rolling median of activity in the time period for the given segment of data, or 3) the nth percentile of the number of entries in the time period for the given segment of data.

The shaded range summarizes information from the data that the control point used to score risk. The top of the range represents the value of the light blue line +1 standard deviation.

Screenshot_showing_a_line_plot_for_the_selected_control_point.png

TrendShock and TrendShock Group-By

When the selected risk score includes a TrendShock or TrendShock Group-By control point, the Entry Details page displays algorithm-specific explainability visuals. These combine visual and textual elements to clarify why an entry is flagged.

  • Scatterplot chart: Shows how the selected entry compares to the broader population, making it easy to see why it stands out.

  • Text explanation: Summarizes why the entry earned its TrendShock control point score, increasing transparency.

  • Control point metadata: Displays the control point name and description from the analysis configuration.

  • Fields of interest: Highlights the data fields used by the algorithm for scoring.

These elements together reduce ambiguity, improve trust in results, and allow auditors and process owners to quickly interpret TrendShock-related anomalies.

image-20250912-192425.png


TrendShock WHERE and GROUP BY support

When a TrendShock control point uses a WHERE filter, a GROUP BY clause, or both, the Entry Details page now reflects the exact population and grouping that TrendShock used during scoring.

This enhancement improves traceability and helps you understand not just that an entry was flagged — but why it was flagged within its filtered or grouped context.

Filtered TrendShock populations (WHERE support)

If the control point includes a WHERE clause, the visuals and tables now show only the entries that were part of the TrendShock scoring population.

  • Entries excluded by the WHERE filter display the message:
    “Not part of TrendShock population”

  • All TrendShock visuals align with the filtered population.

Grouped TrendShock behavior (GROUP BY support)

When TrendShock groups entries (for example, by Brand, Vendor, Business Unit, or Segment), Entry Details now displays the grouped results used for scoring.

This includes:

  • A grouped TrendShock scatterplot reflecting the grouped population

  • A grouped table showing the aggregated values used in the algorithm

  • A related entries table listing the raw entries that make up the selected group

This enables you to move from a grouped anomaly → visual deviation → underlying raw entries.

Combined WHERE + GROUP BY support

For control points that include both filters and grouping:

  • The WHERE filter is applied before grouping

  • Grouped visuals and tables reflect only the filtered population

  • The Related Entries table shows raw entries that match both the group and the WHERE conditions

This ensures alignment between the scatterplot, grouped table, and raw entries.


Export a graph

Select the export icon ( Export icon ) in line with the graph you want to export. The graph will export as a .XLSX file to your computer's default download location.


Work with annotations

Annotations allow you to document and share the insights you have gained from MindBridge. You can add, view, edit and delete annotations and comments on any of the graphs on this page.

Learn how to work with annotations in TRA


Anything else on your mind? Chat with us or submit a request for further assistance.


Related articles

Was this article helpful?