Effective knowledge management for services       

links

 NANDA KAMBHATLA photoKarthik Sankaranarayanan photo

Effective knowledge management for services - overview


Improving productivity of individual practitioners is a critical goal for IT services organizations. In large organizations with many teams providing services to different clients, a critical way to achieve this is by efficiently reusing the knowledge acquired by practitioners.

At IBM Research - India, we are working on developing technologies for effective knowledge management that will enable practitioners to re-use the knowledge in their day-to-day activities. These technologies also provide a general framework for analyzing problem information from IT and application data to automatically identify critical areas for improvement. This requires addressing core research problems in the areas of machine learning, text analytics, information retrieval, and data management. We work closely with the IBM's Services organizations that directly work with our customers.

Some of the technologies under development by our group include

    Knowledge Management Technology for Incident Data:
  • Central web-based repository of problems and resolution history, from multiple accounts, with proper access control
  • Faster analysis of problem by means of Knowledge Reuse: A smart search engine to search for historical tickets for AMS practitioners
  • Quality rating engine information to identify tickets which have useful resolution knowledge and ranks them higher in the search results
  • Protecting data privacy: sanitization of tickets to remove client sensitive information
  • Reduce dependencies on individuals
    Analytics for Identifying Repeat Problem Areas:
  • Identifying candidates for preventive actions in Application or IT services management requires a very detailed profiling of the kind of problems arising in the landscape. Our technology unravels useful information from the unstructured text in the application/IT data to enable teams to get a quick understanding of the patterns of issues. Teams can then dig deeper into specific areas that are causing more grief.Analytics for Identifying Repeat Problem Areas: Identifying candidates for preventive actions in Application or IT services management requires a very detailed profiling of the kind of problems arising in the landscape. Our technology unravels useful information from the unstructured text in the application/IT data to enable teams to get a quick understanding of the patterns of issues. Teams can then dig deeper into specific areas that are causing more grief.

Most existing analytics tools focus on structured data available in the tickets. Recently, there have been significant enhancements in the area of text mining and natural language processing that enable a machine to process free text much more effectively. We leverage these technologies, specifically customized for nature of text in problem tickets, to uncover useful insights for different services teams.