Reference:

Avasthi, V., Dey, S., Jain, K. K., & Mishra, R. (n.d.). The Evolution of Knowledge in Communities of Practice. https://doi.org/10.1145/2811411.2811528

Summary:

  • this paper examines the evolution of knowledge in Communities of Practice
  • in this research article, they establish that the knowledge in a CoP evolves over time – the examine this knowledge, its impact on the community and the individuals involved
  • Section 1: Understanding of Communities
    • Types of knowledge – tacit and explicit
      • Tacit Knowledge – difficult to transfer to another person by writing it down or verbalizing it – it resides in the mind of the person who possess it.
      • Explicit Knowledge – codified knowledge – easily accessed, verbalized, easily transmitted to others, e.g. information from a textbook, the internet, company intranets
    • What are communities
      • they are fluid and may cross boundaries of the organisation, they are not recognized by the organisation
      • as defined by Wenger, communities of practice are mutual engagement in joint enterprise with a shared repertoire
      • they are not self-contained
      • they are groups of people informally bound together by shared expertise and passion common goal
      • in virtual teams, an effective reputation system needs to be a key aspect
    • How are these communities formed
      • knowledge exchange among individuals results in the formation of knowledg networks
      • networks or graphs can be represented by nodes and links or edges
      • community detection within a knowledge graph is accomplished by finding clusters of nodes
  • Section 2: Research Methodology
    • they view Usenet as a collection of multiple CoPs and used its archives as a repository of communication archives of those communities of practice
    • UTZOO Usenet archive¬† is used to analyze the communication patterns across individuals and building of knowledge networks and graphs
    • once the graphs are generated, they are analyzed using network theory and inferences are drawn
    • Topic Identification
      • text mining techniques were used to extract interesting phrases, similarity of texts to enrich the retrieval mechanisms.
      • statistical and machine learning techniques were used to identify and build data in patterns over time
  • Section 3: Results and Analysis
    • they used lingpipe for important phrase connection and igraph for graph generation
    • evolution of communities
      • over time the community change their primary interests
      • there is no gradual increase or decrease in importance
      • once they know the time period for an interesting topic, it will be easy to find an expert in that field from that era
    • evolution of individuals in communities
      • Indegree
        • incoming edges to a node
        • higher indegree signifies higher number of communications where the individual is a receiver – how sought after the individual is in the community
        • indegree of individuals changes over a period of time but once they establish themselves as a sought after individual, they remain so
      • Outdegree
        • outgoing edges from a node
        • higher outdegree signifies higher number of communications where the individual is a sender – how prolific an individual is in the community
        • outdegree of individuals changes over time but once they establish themselves as a prolific contributor, they remain so
      • Betweenness
        • the measure of the individuals centrality in th network
        • higher betweenness signifies higher incidence in an indvidual acting as a go to guy for any two individuals
        • “matchmaker” in a community
        • they may not be the source of knowledge themselves, but they know where to get it
        • betweenness of individuals changes over time but once they establish themselves as a matchmaker, they remain so

(the graphs generated can be seen in the paper)

Advertisements