Telecom companies utilize data mining algorithms and tools to understand thernbehaviors of their customers. Cluster analysis is one of the techniques used tornidentify homogenous groups of customers from a heterogeneous group basedrnon the customersâ€™ service usage records. Clustering algorithms aim to ï¬ndrnnatural groupings of subscribers and are widely applied for customer proï¬lingrnand market segmentation.rnAs telecom customers use services like cellular Internet, huge amount of datarnis generated in the form of call detail record (CDR) primarily for billing purpose.rnThis data can also be a source of information for marketing and networkrnmanagement tasks. Previously, different algorithms have been studied and implementedrnby researchers to understand how, why, when and where peoplernaccess mobile Internet using already available data from the telecom systems.rnIn this thesis, two clustering algorithms namely K-means and Two-Step werernimplemented on real CDR and CRM datasets of a telecom company in Ethiopiarnto segment mobile Internet users. Behavioral segmentation was performed usingrnaggregated data of sample number of customers based on derived features.rnMoreover, the insights obtained from each segment were analyzed before suggestingrnmarketing strategies for personalized services and targeted campaigns.rnK-means was applied on CDR dataset and meaningful clusters showing characteristicsrnof users were obtained and explained. Two-Step clustering was foundrnto be more suitable for segmenting user groups and has a better silhouetternscore than K-means. The results of the analysis show the possibility of usingrndata stored by mobile operators for market segmentation purposes. It wasrnexamined from this research that clustering algorithms like K-means and TwoSteprnare applicable to observe patterns among ethio telecom customers basedrnon service usage and demographic information of subscribers from databasernsystems .