Find out more about setting up Explorer with transactional data in the UI.
So what’s Explorer all about?
Explorer enables users to uncover hidden patterns and identify significant data. To do this,
Explorer examines any flat-file data source, that is, simple data that follows a uniform format and does not contain any complex structures, and represents this data in a variety of formats within the UI.
Why is the feature useful?
Explorer is particularly useful as it enables you to model your data source as an
aggregated Document model, where each Document model represents the aggregation
of several similar Documents. For example, an aggregated transaction Document
depicting 100 distinct transactions between a banking customer and an external supplier.
If there is an aggregated Document model, it is recommended that you have an associated Foreign Document model representing each distinct unaggregated Document. This preserves the data associated with each flat Document that would otherwise be lost during an aggregation, such as the Transaction Date.
Displaying a large table of the unaggregated Foreign Document within the aggregated
Document Viewer is not recommended due to the potentially large volume of Foreign Documents in a typical deployment.
You can use Explorer to run queries over all of the Foreign Document model data, producing useful graphical output such as bar charts, sankey diagrams, word clouds, and geospatial heat maps.
Figure 1. Explorer within the UI
Design considerations
Specific features to consider when using Explorer include:
- The word cloud graph is a powerful graph that can be applied to free text fields. However,
deployments have reported issues with performance over large volumes of free text. Graphs can be configured using the deferLoad property to only load under certain conditions. This can help with performance issues associated with graphs loading too many results. - Prefilters are useful when regularly performing the same queries. Construct Explorer queries using a URL for repeated execution. Additional information can be found within the Prefiltered queries documentation.
- Significance aggregations are useful for comparing a subset of data against a larger set of data to detect anomalies. For example, Chemical Compound A for creating explosives appears in 1% of all transactions, but when compared to transactions involving Company X it goes up to 4%, making it a significant anomaly.
Additional information
For additional information on this topic, refer to the Explorer documentation.
Also, if you have any ideas for Explorer you can submit them via our Ideas Portal.
Build information
Version 2.2.0, Dec 2022
