Airlines ATA mapping Automation using Text Mining
3AI November 4, 2020
Airlines based in United States of America
Problem Statement
The airline inspection is done on paper and the engineer is tasked to write down the issue found and assign appropriate ATA code.
There are 2000+ ATA code codes and the engineers do not remember them. Therefore write any random ATA code against the issue. The challenge is to identify and assign the correct ATA code to each entry
Solution Approach
-
Warehouse Server
-
Word Frequency Algorithm
-
Create Taxonomy
-
Scoring Model
-
Real-time Processing
Business Impact
-
60% reduction in the digitization process and therefore the reduced digitization cost
-
Revenue earning opportunity for client to monetize this as they manage engineers for other airlines in their network
Critical Success Factors
- The model gives top 5 score ATA against Text. The top most is correct mapping
- In case of ambiguity the remaining scored ATA code can be looked at
