Structuring Cross-Organizational Knowledge Sharing

Ontology development is fundamental to most knowledge management efforts. When approached in a formal knowledge engineering manner the resulting ontology usually becomes brittle when spanning even a modest number of groups within a single organization. It breaks entirely when scaled to multiple, heterogeneous organizations. A promising alternative is the bottom-up approach such as can be found in social tagging systems (e.g., del.ico.us), but little research has examined the utility of these systems for knowledge reuse activities. In this paper we extend our field work with IT helpdesk staff to examine the drivers for natural ontology development. We found that a balance between some degree of external order while maintaining local flexibility was required. This information space is navigated via social relations, especially expert referral. We examine the user-centered design criteria for both mid-level ontology development and related expert profile management.