/Airline-Data-Warehouse-Modeling

Analyze a major airline company's current business processes and expanding the company by discovering new opportunities by Analyzing flight activities, reservation process and customer care interaction.

Airline-Data-Warehouse-Modeling

Analyze a major airline company's current business processes and expanding the company by discovering new opportunities by Analyzing flight activities, reservation process and customer care interaction.

Applied using Bill Inmon data model, ERD- Dimensional Modeling (snowflake schema)

The Grain:

Flights per segment (Transit). Segment is the stop point for each flight if it’s not a one way flight

Dimensions:

Passenger, Flight, Promotion, Flight_Status, Resv_Channel, Feedback, Feedback type, Class_type, FareBasis, Airport

Facts:

  • Flight_activity: Total_miles, Segment_Duration,ticket fees.
  • Promotions_respose: Factless table(Passenger_key, Promotion_key, Flight_key)
  • Class Upgrades: Factless table(Class_key, Passenger_key, Flight_key)
  • Passenger_Interactions: Total complaint, Total feedbacks, total inquires.

Assumptions:

  • To be a frequent flier you must exceed 6000 miles, (the total miles for that passenger flight)
  • Passengers often Upgrade, Downgrade or don’t change their class type for the flight 11
  • Passengers is a (Slowly Changed Dimension) to change the passenger_flight_status when they earn a certain total amount of miles. If total mile more than 5000 and less than 10000 then he will have titanium status else if total mile more than 10000 and less than 5000 he will be upgraded to platinum, Else if he exceeded 10000 miles he will have a gold status.
  • When Receiving feedbacks from passengers, it’s divided into 3 categories, either it’s a problem, inquiry or a feedback, and if it’s a problem it will have an indicator problem severity, which have a rate from 1 to 10 to indicate how severe is that problem