Mobile phone data can provide rich information on human activities and their social relationships which are dynamic in nature. Analysis of such social networks emerging from phone calls of mobile users can be useful in many aspects. In this paper we report the methods and results from a case study on the analysis of a social network from mobile phone data. The analysis involves tracking the dynamics of the network, identifying key individuals and their close associates,
and identifying individuals having communication pattern similar to the key individuals. We introduce novel measures to quantify, the evolution in the network, significance of an individual, and social association of an individual. In order to
group individuals having similar communication pattern, we applied recently proposed online clustering approach called eClustering (evolving clustering) due to its adaptive nature and low computational overhead. The results show the pertinence of the proposed quantification measures to analysis of evolving social network.