Knowledge management with all its aspects of data gathering, knowledge construction and representation, and the facets of human interaction with knowledge systems has been a very active field of research and development over many years. A very detailed description of the theoretical foundation, the processes and systems involved, and the organizational challenges of knowledge management can be found in [3]. Here, we restrict ourselves to present a selected list of references that influence and complement our work. At its core, our work follows the three phases of the framework for strategic knowledge management as described in [4]. In the knowledge creation phase, we inspect available systems, data sources, and processes to explore hidden and tacit knowledge. In the transfer phase, we observe professionals to transfer their implicit use of knowledge into our data structure. In the utilization phase, finally
we support and encourage the application and extension of knowledge by the professionals in the field. As outlined in [5], our approach embeds technology into a knowledge management process to address the organizational dynamics in a services transition environment. Knowledge transfer is one of the core objectives in services transition. In [6], the authors have postulated the importance of face-to-face conversations and side-by-side cooperation for efficient knowledge transfer. The additional challenges arriving from geographical distances and cultural differences are described in [7]. Our goal therefore has been to provide adequate technology to utilize the limited face-toface time to its fullest extent and to encourage agile knowledge sharing between on-shore and off-shore resources. While knowledge sharing seems obviously of great benefit to an organization at large, there might be personal motivations to not share or accept knowledge. Among such motivations are mistrust in peers and management, fear of losing a professional advantage, and belief in the individual’s superiority [8]. Therefore, an incentive system is needed to encourage knowledge sharing. In [3], three basic classes of incentives are listed: (1) financial incentives or some form of material reward, (2) moral incentives or acknowledgement expressed though awards and status, and (3) threat of punishment in case of failure to act. While our approach does not directly implement a particular incentive system, it provides the data and functionality to support such a system in accordance with the requirements listed in [9], such as transparency, flexibility, performance, and economy of incentives. Our data model for knowledge representation and the core of our underlying system architecture are based on our prior work on Business Provenance [10]. In a nutshell, business provenance records the entities of business operations – such as tasks, documents, actors, processes, policies, milestones, etc. – and their manifold relationships in a semantic network. Semantic networks are a widely used mechanism. Similar data models can be found for instance in [11] and [12]. But business provenance is not just a data model. Moreover, the extensible provenance engine constantly analyzes and correlates the incoming data to derive new insight. In the approach presented in this paper, we have added new functionality to the provenance engine; to classify entities based on text mining [13], [14]; and to discover correlations between entities automatically [15], [16]. A smart way of data gathering, a sophisticated knowledge representation, and an advanced set of analytics mechanisms are necessary ingredients to an effective knowledge management system in AMS operations. However, the ease of use, the precision of answers, and the value of recommendations are those factors that finally determine the acceptance and thus success of the overall effort. To address those needs, we incorporate search mechanisms that are based on the ‘social network’ of each user [17]. By doing so, we are able to present search results ranked by familiarity and similarity as well as relevance to the history of the user or the topic at hand. Furthermore, the user’s ‘social network’ is the basis to determine her/his expertise – both through the authoring and the consumption of knowledge [18]. Thus, our system is able to provide recommendation on ‘Who knows what’. Within the incentive system that encourages sharing of knowledge, identified experts act as knowledge multipliers within the organization.