Serving Information Needs in Business Process Consulting

摘要

Business Process Consulting is a knowledge-intensive activity that requires consultants to be aware of all available process variants and best practices to implement the most effective business transformation. Typically, during a business-consulting engagement, a large amount of unstructured documentation is produced that captures not only the processes implemented, but also information such as the requirements addressed, the business benefits realized, and the organizational changes impacted. Consulting organizations archive these documents in internal knowledge repositories to leverage the information in future engagements. Our analysis of one such repository in the IBM Global Services' SAP practice indicates that although there is tremendous scope for information reuse, the true potential of reuse is not being fully realized using existing search technology. We present a novel search technique to serve the information needs of consultants as they script various pieces of design information. The proposed system auto-formulates fat semantic queries based on the context of the design activity and runs the queries on a semi-structured knowledge repository to return a ranked set of search results. Our empirical studies show that our approach can outperform-often, significantly--a conventional keyword-based search technique in terms of both relevance and precision of search results.