Cloud Computing: Knowledge Management

Knowledge Management in a cloud computing environment requires a paradigm shift, not just in technology and operational procedures and processes, but also in how providers and consumers of knowledge within the enterprise think about knowledge. The knowledge-as-a-service, “on-demand knowledge management” model provided by the cloud computing environment can enable several important shifts in the way knowledge is created, harvested, represented and consumed.

Collective intelligence is a phenomenon that emerges from the interaction – social, collaborative, competitive – of many individuals. By some estimates there are more than eighty million people worldwide writing web logs (“blogs”). The blogs are typically topic-oriented, and some attract important readership. Authors range from large company CEOs to administrative assistants and young children. When taken together, the cloud computing infrastructure which hosts “blogospheres” is a big social agglomeration providing a kind of collective intelligence. But it is not just blogs that form the collective intelligence – the phenomenon of collective intelligence is nurtured and enhanced by the social and participatory culture of the internet, so all content developed and shared on the internet becomes part of the collective intelligence. The internet then, and the content available there, appears as an omnipresent, omniscient, giant infrastructure – as a new form of knowledge management. This same paradigm applies, albeit on a smaller scale, to the enterprise cloud – the socialisation and participatory culture of today’s internet is mirrored in the microcosm of the enterprise.

Today this represents collaboration of mostly people only, but very soon in the future we may envisage intelligent virtual objects and devices collaborating with people. Indeed, this is already beginning to happen to some extent with internetattached devices starting to proliferate. Thus, rescaling from the actual ~1.2 billion users to tens or even hundreds of billions of real-world objects having a data representation in the virtual world is probably realistic. It is important to note here that content will no longer be located almost solely in a central knowledge repository on a server in the enterprise data centre. Knowledge in the cloud is very much distributed throughout the cloud and not always residing in structured repositories with well-known query mechanisms.

Architectural View Of Enterprise Knowledge Management

Knowledge management applications offered in the cloud need to be capable of crawling through the various structured and ad-hoc repositories – some perhaps even transient – to find and extract or index knowledge, and that requires those applications be capable of recognising knowledge that might be useful to the enterprise knowledge consumers. Furthermore, we believe that over time multimedia content will become dominant over ordinary text, and that new methods for media-rich knowledge management will need to be devised. Even in the smaller world of the enterprise, a real danger, and a real problem to be solved by knowledge management practitioners, is how to sort the wheat from the chaff – or the knowledge from the data and information – in an environment where the sheer amount of data and information could be overwhelming. The best domain for Enterprise Knowledge Management is in the Enterprise IT domain, as it is a domain under huge cost pressure but one which is essential for strategic development. From a highly abstracted view, the Enterprise Knowledge Management IT domain consists of problem solving, monitoring, tuning and automation, business intelligence & reporting, and decision making tasks.

The tasks of problem solving, monitoring, tuning and automation, business intelligence and reporting, and decision making are the most promising areas for the future deployment of Enterprise Knowledge Clouds. The knowledge available to both IT administrators and automated management agents via the Enterprise Knowledge Cloud will help drive the development of a slew of new technologies addressing the problems which previous computing facilities couldn’t resolve. Currently, the majority of the indicated IT tasks include people, while we suggest that this balance will be changed in the future through automation, ultimately leading to self-managing enterprise IT systems. When mapped into more precise form, this conceptual drawing will evolve into the enterprise-scale knowledge management application stack.