Customer story · Golden
Incomplete travel data.
Two usable masters.
IT Travel Services needed reliable references for the people travelling and the places they visited. Golden was used to resolve incomplete traveller records and build a destination catalogue, giving travel reports a consistent basis for identifying people and places.
Travel technology · Master data
- Traveller source records
- 5M+
- Destinations in the catalogue
- 250K
Names and places
need interpretation.
The traveller data offered limited information: names and surnames, often incomplete, expressed as free text. Before the records could support useful reporting, variations had to be normalised and duplicates resolved. With so few attributes available, identity resolution depended on how that information was interpreted.
Destinations presented a different difficulty. A free-text query might refer to a city, an airport, a station or a point of interest. Making it useful required a catalogue of places to search against and a way to find matches when the wording did not coincide exactly.
Two master domains
Different inputs.
A reference for each.
Golden addresses the two domains separately: normalising and deduplicating traveller records, and combining an open-data destination catalogue with approximate search.
Traveller input
Incomplete names
Names and surnames supplied as free text, with limited information for identifying the traveller.
Trazadera Golden
Resolve identities
Normalise name variations
Identify duplicate records
Consolidate the traveller master
Traveller master
Traveler Finder
Reporting that uses a common traveller reference across the source records.
Destination input
Places & free text
Open data describing cities, airports, stations and points of interest, alongside free-text destination queries.
Trazadera Golden
Build & match
Create a world destination catalogue
Search with approximate matching
Resolve queries against the catalogue
Destination master
A common place reference
Analytical reports use master destinations to relate travel activity to the places it concerns.
What changed
Master data behind
the reports.
The project produced two distinct reference datasets. Traveller resolution supports Traveler Finder; the destination catalogue supplies a common location reference for analytical models. Each tackles a different ambiguity in the original information.
Mar Muñoz describes the result in terms of the value of IT Travel’s analytical reports. The work sits beneath those reports: deciding how incomplete names and variable destination text relate to the people and places being analysed.
Which reference data
still arrives as free text?
Bring a few representative records and the reports or systems that need to make sense of them.
Talk to Trazadera