Telecom analyticsday
Download as PPTX, PDF1 like225 views
This document discusses various methods for detecting fraud in the telecommunications industry, including subscriber, recharge voucher, prepaid balance, unauthorized service, and SIM cloning fraud. It also outlines challenges with churn faced by a telecom provider called Bad-Idea and potential solutions around location-based services and a dating referral service. Finally, it provides assignments for students around fraud detection, churn reduction, and campaign management involving building presentations, data models, surveys and SQL queries.
Telecom analyticsday
- 1. Telecom Analytics Day III
- 2. Detecting Subscriber Fraud . . . • High number of calls to Black Listed numbers • High Roaming charges • High Internet Usages • High number of VAS calls • Frequent Change of Address • Pre-Subscription Check: • Verify address • Verify home number • Set Credit Limits • Check PAN number, UID against Credit Violations • Check IMEI against Black Listed IMEI • Check for matching names with black listed customers. • Check for matching PIN codes. • Check for addresses from notorious localities. • Match subscriber usage profile with black listed subscribers : • Called numbers • Matching tower locations • Calling patterns (short calls, long calls)
- 3. Detecting Recharge Voucher Fraud . . . • Unusual top-ups • High number of recharges in a given time-period
- 4. Detecting Pre-paid Balance Fraud . . . • Track employees with high number of manual balance change • Subscribers with high balances
- 5. Detecting Unauthorized Service Fraud . . . • HLR vs. Postpaid subscriber profile reconciliation • HLR services vs Postpaid Subscriber services Profile mis-match • Sudden change in Subscriber usages (??)
- 6. Detecting SIM Cloning . . . • Velocity Check • Call Collision
- 9. Churn problem at Bad-Idea. . . The CEO of Bad-Idea has come back Praxair Analytics Inc. with a new problem. He was very impressed with the fraud solution implemented by Praxair and hopes that they will be able to help him out. Over the last two years after Mobile number portability was introduced, about 20 million subscribers has become inactive or has left Bad-Idea(post-paid users).
- 10. Bad-Idea… Let’s open a dating service The CEO of Bad-Idea is under a lot of pressure from the board to increase his ARPU. The marketing team has come up with a set of ideas. One of this idea is to start a service similar to Shadi.com. He plans to start with sending a trial offer of the current subscriber base. He does not want to spam all the subscribers with this offer. He wants to find out the subscribers who will be interested in this service.
- 11. Assignment Fraud Detection (20) . . . 1. Make a presentation to the CEO of Bad-Idea about how you can help him with Fraud detection. • Introduction to the company (make your imagination go wild) • Typical Fraud Scenarios • Case studies of companies who benefitted from your solution • How you differentiate from the competitors. 2. Design the Star Schema for fraud analysis. 3. Collect at least 25 phone bills for the same month. Load this data into a database as a fact table. 4. Write SQL queries to detect fraudulent conditions. (You will get results if your thresholds are unrealistically low). 5. Deliverables: • Power point presentation • ER Diagram of the star schema • SQL queries and output. • Project Plan
- 12. Assignment Churn Reduction (20) . . . 1. Make a presentation to the CEO of Bad-Idea about how you can help him with Churn Reduction. • Introduction to the company (make your imagination go wild) • Typical Churn Scenarios • Case studies of companies who benefitted from your solution • How you differentiate from the competitors. 2. Create a survey and ask your friends and families to complete it. The survey should capture instances when the had a bad experience with their telecom provider. The survey should also capture what will it take for these people to switch telecom providers. (Collect at least 25 surveys) 3. List how you can detect these churn scenarios along with the subscribers. 4. Design the Star Schema for detecting Churn candidates. 5. Write Pseudo SQL queries to detect these conditions. 6. Deliverables: • Power point presentation • ER Diagram of the star schema • Survey Questions • Survey results in a quantifiable manner • SQL queries • Project Plan
- 13. Assignment Campaign Management (20) . . . 1. Make a presentation to the CEO of Bad-Idea about how you can help him with Location based services. • Introduction to the company (make your imagination go wild) • Example of location based services • Case studies of companies who benefitted from your solution • How you differentiate from the competitors. 1. List at least 20 businesses that will benefit from Location Based Services. (Hint: Stuff we buy on impulse..). List 10 businesses that will not benefit from Location Based Services. 2. Identify the potential profiles of a person that will probably be influenced by the message for each of these selected services. 3. Design the Star Schema for detecting the profiles. 4. Write Pseudo SQL queries to detect these conditions. 5. Create a survey and ask your friends and families to complete it. Ask them whether they will be influenced by the location based offers. Also capture relevant demographic information. For each service category list the following: • False Positives (People who will receive the offer and will not take it up) • False Negatives (People who would have taken this offer but were filtered out based on your criteria) • Positives • Negatives. 6. Deliverables: • Power point presentation • ER Diagram of the star schema • SQL queries • Survey Questions • Survey results in a quantifiable manner • Project Plan
Editor's Notes
- #3: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #4: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #5: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #6: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #7: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #8: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #10: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #11: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #12: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #13: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.
- #14: Why we had the second UAT. Impression of ICTEAS not being serious, cause we should be able to generate the right report, right accountability.