Semi-Supervised Algorithm to Detect Over-the-top Bypass Fraud: In the Case of Ethio Telecom
| dc.contributor.advisor | Yihenew, Wondie (PhD) | |
| dc.contributor.author | Wubalem, Kinfmichael | |
| dc.date.accessioned | 2022-02-09T07:40:01Z | |
| dc.date.accessioned | 2023-11-04T15:13:03Z | |
| dc.date.available | 2022-02-09T07:40:01Z | |
| dc.date.available | 2023-11-04T15:13:03Z | |
| dc.date.issued | 2021-11 | |
| dc.description.abstract | Telecom fraudsters' behavior evolves over time, and their capacity to defraud telecom service providers grows at a rapid pace. Interconnected bypass fraud is one of the most common types of telecom fraud. It is a new sort of telecom fraud that works by intercepting international voice calls and transferring them to VoIP for termination as an OTT call without the knowledge of the telecom provider, caller, or called party. OTT frauds based on the Mobile Station International Subscriber Directory Number (MSISDN) are becoming more common, posing a threat to telecom firms as the number of smart phones grows and it becomes easier to access OTT services from anywhere. Interconnected bypass fraud is the term for such operations, and OTT bypass is one sort of interconnected bypass fraud. One subtask in the detection of this scam is detecting OTT voice call packets using various network traffic classification techniques. To categorize network traffic packets, machine learning (ML) techniques are utilized, including the semi-supervised and supervised algorithms Collective Filtered and Decision Tree (DT). For the training and testing of the chosen method, ten cross-fold validation techniques were used. Each ML algorithm's test dataset is correctly prepared. With reasonable model build and evaluation times, both ML algorithms DT and collective filtering achieve better performance of 99 percent and 91 percent accuracy, respectively. | en_US |
| dc.identifier.uri | http://etd.aau.edu.et/handle/123456789/29972 | |
| dc.language.iso | en_US | en_US |
| dc.publisher | Addis Ababa University | en_US |
| dc.subject | OTT bypass fraud | en_US |
| dc.subject | Telecom fraud | en_US |
| dc.subject | Network traffic classification | en_US |
| dc.subject | Machine-learning algorithms | en_US |
| dc.title | Semi-Supervised Algorithm to Detect Over-the-top Bypass Fraud: In the Case of Ethio Telecom | en_US |
| dc.type | Thesis | en_US |