I’m in Denver this week attending the Association of Library and Information Science Educators iCreate Conference 2009.
Tuesday will be a busy day as I’m participating in the WISE workshop panel:
Stepping out of CMS: Student Communication Technologies Beyond the Course Management System
Panel Presentation and Discussion of effective practices for instructor/student, student/student, and student/instructor communication strategies outside the context of the online course management system, as well as with the wider community of LIS professionals, alumni and prospective students.
and I’ll be presenting a “work in progress” poster:
Blended Methods for Ontology Development
Ontologies represent an important backbone for knowledge representation on the emerging Semantic Web. As a formal specification of concepts within a domain, developing an ontology requires translating the knowledge of domain experts into the classes, properties and relationships used by machine-processable languages such as RDF and OWL. Current ontology development practices owe much to knowledge and software engingeering processes, however the methods for capturing the knowledge of domain experts reamins under-theorized. While “mixed” qualitative and quantitative methods have received extensive discussion in the literature, less attention has been paid to blending the kinds of formal methods used in ontology development and qualitative methods used elsewhere in LIS research. The resulting “knowledge acquisition bottleneck” has lead some ontology developers to turn towards mining large textual datasets for base concepts using natural language processing techniques. While these tools are improving, automated population of an ontology still requires intervention and evaluation by domain experts – particularly in areas where textual sources present conflicting or incomplete representations of a domain.
Lee (2000) has identified the lack of agreement on concepts of “collections” among LIS professionals and their users – exactly the kind of domain that challenges automated techniques. The research discussed here is working towards an ontology for cultural heritage “collections” as identifiable entities that are more than the sum of their parts. As part of the work in progress, this poster explores how qualitative approaches, such as Glaser & Strauss’ Grounded Theory, can be used to inform the development of such an ontology.
Of course this “work in progress” abstract was written a few months ago and after digging into this topic a little deeper the focus of my poster has shifted just a bit. What you’ll see tomorrow (I’ll post the final version here) focuses more on the similarities between Strauss & Corbin’s open/axial coding process and ontology development. As I mentioned in my last post, I’ve backed away from developing an ontology for its own sake and the revisions to the poster reflect my current thinking about how ontologies could inform traditional QDA approaches. Along these lines, the poster also explores the possibility of using the CIDOC-CRM (or any existing ontology) as start-list of qualitative coding concepts.
Stay tuned for Twitter updates!