It could be argued that stating "socialisation is the key to knowledge sharing" is so implicit that it should not even be talked about it. But let's talk about it, and especially from a technical perspective.
Knowledge sharing is all about transferring knowledge from one person to another, or about the second person knowing that another person has some knowledge which they might eventually need. This has been the basis of various Knowledge Management models and concepts including the SECI model (Nonaka, 1994), Communities of Practice (Wegner) etc. The key element in these models and techniques, and others similar to them, is that knowledge is shared socially, i.e. between people.
IT systems have been implemented to provide solutions into helping Knowledge Management (KM). This is not surprising at all as IT fits in most aspects of our lives nowadays. Undoubtedly, IT has increased work efficiency in a direct or indirect way for every job that exists nowadays. However, in terms of IT applications for KM, IT has yielded a mix of results in terms of success.
As written in a paper I co-authored, socialisation seems to be have been disregarded when implementing IT-based Knowledge Management Systems (i.e. IT solutions to help a knowledge management initiative). Taking the technical perspective, for example, this seeks to implement IT as a solution to knowledge management problems. On the other hand, the socio-technical perspective, argues that people are important to be considered as part of the knowledge management system (KMS). With the argument that knowledge can only exist within humans, it seems natural that humans need to be mostly involved within a system (hence aligning mostly with the socio-technical perspective).
However, it seems that IT-based solutions have always put an extra layer of interface between the humans in the system, possibly not really simplifying the knowledge sharing process. Although this seems to be inevitable in order to facilitate knowledge indexing, and its reuse, these systems seem to lack into providing any extra benefits. Research suggest that humans are very selfish, and they only do something, if the returns of participation are higher than the costs. For example, the effort of sharing a piece of knowledge to a repository is perceived as cost of time, plus a knowledge cost too. Therefore an employee would expect something in return in order to actively participate. If not many people actively participate (by sharing) then the returns (new knowledge from others) from the system are very low, and the knowledge within the system risks to become stale and outdated very quickly.
Am I suggesting a KMS without IT? No, this would definitely be a step backwards, especially for repositories and their management. The key word is socialisation. How can IT help in socialisation, and in turn, help in KM? The answer might lie in social networking technology.
As Granovetter (1973) points out, social networks are great for sharing and acquiring new knowledge. Especially connecting with 'weak ties', i.e. people who have a very different network to ours, has the advantage of bringing new knowledge into our network. The more the sources, and the more diverse they are, the bigger the knowledge potential. A network that spans wider, will hold more interesting information and knowledge. The phrase of "it is not what you know, but who you know" is almost entrenched in how most of the world works.
From a technical perspective, the potential to socialise, connect and follow people (their actions, their work, and ultimately their knowledge) is nowadays further simplified through the aggregate use of Web 2.0 technologies, under the guise of public social networks (such as facebook, twitter, linkedin etc.) Knowledge sharing is clearly happening through these networks (link here).
In the past year, I have been involved in putting all the above into practice for a KM initiative within a company that was willing to participate in the study. The results, yet to be published, are astounding, albeit not much of a surprise. Connecting people together, helps span geographical boundaries, and allows ad-hoc knowledge seekers to be able to easily connecting with knowledge owners. Morevore they allow knowledge to find the right recipients (through follows).
Socialisation is the key to sharing knowledge, whether enabled by a process, events, or IT. KM should seek to connect people together and enable to share knowledge in more efficient ways.
(References to statements can provided upon request, I just didn't want to turn this into a copiable academic paper - which is in the process of being published)