Monitoring for Anomalies in Data Segments

Now that your data is ingested and ready, you can start configuring Mona's insights generator.

Mona's Insights Generator's job is to find specific segments in your data on which the data behaves in an anomalous manner, according to your preferences. This is used to find drifts, biases, data integrity issues, and more, even when those occur only in specific places in the data.

Mona also takes a smart approach to noise reduction, making sure that if an anomaly can be found in several ways, you would only get alerted once with the full information of what's happening in the data.

Before starting, let's review the main concepts in Mona's Insights Generator.

Verses:
Verse objects are the configuration atoms of Mona’s anomaly detection engine.
Each Verse contains instructions to detect a single anomaly type within the monitoring context. These instructions specify data segmentation configurations, the metrics to measure and compare, and other specifications regarding assessing whether the behavior is anomalous.

Stanzas:
Stanza objects contain a list of Verse objects, typically ones that share some common parameters. The intent behind Stanza objects is to logically group verses with shared parameters under one name and enable you to define these shared parameters once. The top-level key is the name of the Stanza object.

More information on verses and stanzas can be found here.

The following chapters will show you how to set up three basic anomaly detection verses - AverageDrift, AverageOutlier, and AverageSuddenChange.