Journey Discovery Workflow
- Pete Trevitt
This page gives an overview of the Journey Discovery execution process with a view to making sure Analysts understand the different elements and when and how they can be used.
Background Knowledge
A Journey Discovery execution is built up thought 4 key dependant stages, each relying on all those before it.
These stages are as follows;
Stage 1) Selection and Saving of Data
Select Event Stream
Validate Data Transform - Ensure a comprehensive Data Transforms is in place or results will be of no value.
NOTE: Data Transform version will be tied to this Journey Discovery process and cannot be edited. Any changes to the Data Transform within that application, will not be inherited here.
Validate Location Map - Create as best as known, a Locution Map which any Known Journeys.
NOTE: Data Transform version will be tied to this Journey Discovery processed and cannot be edited. Any changes to the Data Transform within that application, will not be inherited here.
Select Date Range
Save and Calculate Data (first execution)
Date Range is selected and Creates/Saves the resultant data set that the Journey Discovery process will be executed on. This is a fixed data date and will persist for the life of this Journey Discovery.
Immediately creating a new Date Range now, would simply over-write
At this stage the Analyst can only move onto Stage 2
Stage 2) Journey Discovery Processing
In stage 2, the saved data from stage 1 is passed into the Data Mining function to create the Micro Journeys that exist in the date. Theory changing ay configurations if Stage 1, would mean that this process would also need to be repeated.
If the Analyst goes back to Stage 1, changes any configuration and re-runs there, this will invalidate the Stage 2 data here from the previous data set.
The Analyst can move onto Stage 3
Stage 3) Existing Journey Ranking
At this stage, the Analyst is simply looking at a list of Journeys that have already been mapped in Location Mapping, and ranking them in order of “Analytical Importance”. This is a a way of ensuring that patterns from Stage 2, that contains Locations from a Ranked Journey, are flagged as higher importance for future analytics.
Save and Calculate Data (first execution)
Executes the Journey Discovery Ranking process, built upon the Micro Journey output from Stage 2,based on the ranking completed in Stage 3
Displays the Discovery Result For Stage 4 review
If the Analyst goes back to Stage 1, changes any configuration and re-runs there, this will invalidate the Stage 2 data here from the previous data set, and invalidate the result as the data set will have changed.
If the Analyst goes back to Stage 2 , changes any configuration and re-runs there, this will invalidate the Stage 3 data for ranking and invalidate the result as the Micro Journeys will have changed.
Stage 4) Discovery Result
Result Review screen