Corresponding Author

Pradeep Kumar, ICFAI Business School, ICFAI University, India (

Document Type



Successful e-commerce portal organizations focus intensely on customers. They try to consider every bit of information that flows from the customer to their system as an input for analyzing and identifying their needs precisely and catering to them. Being mostly ‘click and mortar’ or completely e-enabled, they have a lot of operational flexibility to address customer requirements in a more personalized and customized way than their brick-andmortar counterparts with more operational rigidity and resource constraints.

Managing diverse range of channels is a challenge because of exponential and sometimes disruptive growth of diverse technologies that are used for supporting highvolume e-commerce operations. Customers are bouncing between phone, email and the web with greater fluidity than ever and therefore, fragmented, ‘stove-pipe’ communications, in such situations, can create problems as they loose out the holistic view on the basic nature of the problems and customer priorities. Therefore, the use of a common knowledge base across all channels is a dire necessity for an e-commerce portal, especially the ones which do not have a ‘brick-and-mortar’ back-end. The customer-support knowledge network as proposed in this paper addresses these issues. Using Self Organizing Maps(SOM), the network becomes incrementally self learning representing various groups of communication instances at any point of time. The advantages include the integration of all communication elements and an assimilation of all the customer communication issues into a reusable form of self-learning network. It adds an immense value for a customer-focused ecommerce company for identification of generic issues, better understanding of customer concerns and priorities and designing products/ services/ promotions accordingly, to ensure an overall better success of business.