Business Context
An IT department wants to automatically categorize support tickets into 'Hardware' or 'Software' categories based on keywords found in the ticket descriptions to route them to the correct support team efficiently.
About the Set : boolRule
Extraction of Boolean rules for classification.
Discover all actions of boolRule Data Preparation
Creation of a small dataset representing IT tickets. Terms are mapped to IDs (e.g., 10='Screen', 11='Crash').
Copied!
| 1 | |
| 2 | DATA casuser.ticket_terms; |
| 3 | INPUT ticket_id keyword_id; |
| 4 | DATALINES; |
| 5 | 1 10 1 12 2 10 3 15 4 11 4 15 5 11 6 12 ; |
| 6 | |
| 7 | RUN; |
| 8 | |
| 9 | DATA casuser.ticket_info; |
| 10 | LENGTH category $10; |
| 11 | INPUT ticket_id category $; |
| 12 | DATALINES; |
| 13 | 1 Hardware 2 Hardware 3 Hardware 4 Software 5 Software 6 Software ; |
| 14 | |
| 15 | RUN; |
| 16 |
Étapes de réalisation
1
Load data tables into CAS (Implicit in data prep, verifying existence)
Copied!
| 1 | |
| 2 | PROC CAS; |
| 3 | TABLE.tableExists RESULT=r STATUS=s / name='ticket_terms'; |
| 4 | |
| 5 | RUN; |
| 6 |
2
Execute brTrain to generate boolean rules for classification
Copied!
| 1 | |
| 2 | PROC CAS; |
| 3 | boolRule.brTrain / TABLE={name='ticket_terms'} docId='ticket_id' termId='keyword_id' docInfo={TABLE={name='ticket_info'}, id='ticket_id', targets={'category'}} casOut={name='ticket_rules', replace=true}; |
| 4 | |
| 5 | RUN; |
| 6 |
Expected Result
The action successfully generates a 'ticket_rules' table containing IF-THEN logic linking terms (like 10/Screen) to the 'Hardware' category and terms (like 11/Crash) to 'Software'.