Proceedings of 6 th International Conference on Education and Information Management (ICEIM-2014) 97 A Framework for Mining Top Call Drivers as Experienced in Technical Contact Centers to Facilitate Self Help Portals *Ashish Dutt 1 , Saeed Aghabozrgi 2 Faculty of Computer Science & Information Technology (FSKTM), University of Malaya Lembah Pantai, 50603 Kuala Lumpur, Malaysia *ashish_dutt@siswa.um.edu.my Abstract: The presence of information technology has become an intricate part of our life. Few would have realized in the early eighties that in the near future there will be multi-million dollar service based industries. The industrial revolution was successful because man knew how to work with machines. For the present knowledge age revolution to be successful it will depend on the wisdom inscribed in knowledge base articles. The absence of relevant knowledge base articles in self-help portal of technical contact center results in high customer call volumes in service industry. This problem can cause staffing issues as well as loss of business credibility. That can further lead to staff attrition rate and business loss. To resolve this issue we propose a framework. We have used statistical methods and computer programming language to quantify the various hypotheses central to developing the proposed framework. Web based questionnaire and interview are used to collect data for this research. In this research an interesting finding was the null hypotheses that Average call Handle Time (AHT) and Knowledge Base (KB) referral were not related to each other, and that is a contentious paradigm. This is reasoned to less data collected for research evaluation. Keywords: Data mining, Self Help Portals, Java, Python, XML 1.Introduction As technology advances it has impending problems associated to it. One such notable example is a service center also dubbed as the contact center for technology viable products like a computer or a mobile device or a television set. While technology strives toward achieving ubiquitous computing, so is the consistent need for repairing or patching its components too. However, to repair any electro-mechanical device, it is important to know the correct solution but generally these solutions are distributed unevenly across various mediums. Examples of these mediums could be varied either in print form or non-print form, they range from books, articles, newspapers, hardware, and software and off recently a very business oriented solution has sprung up and that is Dzself-help portalsdz. (owever the efficacy of these self-help portals in most cases is not relevant at times. Most of the technological companies provide self- help portals along with their contact centers to help consumers resolve any technical difficulty with their products. When a consumer contacts the technical helpdesk their complaint is recorded in a call log. These call log contain both the problem description as well as the solution provided to the consumer to solve the problem. Besides the technical helpdesk, most Information Technology based companies also provide web based help too. These are also known as the self-help portal. Much can be learned from the call logs of these contact centers to help improve the quality of the self-help portals by integrating the top call drivers of a contact center with the knowledge base of the self-help portals, which could then lead to improved customer satisfaction. This will also help to reduce the operational costs of the contact centers. There exists a knowledge gap between knowledge base of self-help portals and the solution sought by the consumer. Another issue is the presence of obsolete or outdated information within these self-help portals because their knowledge bases are seldom updated regularly. Yet another issue is the business perspective, wherein at times some of the self-help portals only present parts of desired information. Reason for this is that sometimes these solutions are too technical in nature for a layman to follow. To summarize, the absence of relevant knowledge base articles in self-help portal of technical contact center causes high customer call volumes. This problem can cause staffing issues as well as loss of business credibility. That can further lead to staff attrition rate as well as business loss. It is realized that the outsourcing industry requires a framework that could effectively employ and use the indirect knowledge generated by its knowledge workers. A contact center agent uses Knowledge Base (KB) article that help them to resolve consumer queries. During this process it inadvertently also generates indirect knowledge. Sometimes the technical contact center support agent develops quicker resolution steps than the ones described in the self-help portal which they write in the customer call log. This helps in quicker customer call resolutions and better business. This indirect knowledge is an asset for the company and if it is not tapped then it results in either be a business or people loss. In this study, we propose data mining techniques to mine the call logs. We have used data mining methods that can help consolidate this uneven spread of knowledge components in technical contact center consumer call