Transitioning Existing Content: inferring organisation-specific documents

  • Arijit Sengupta
  • Sandeep Purao
Keywords: transition, organisational data, method, heuristic, artificial intelligence, natural language


A definition for a document type within an organization represents an organizational norm about the way the organizational actors represent products and supporting evidence of organizational processes. Generating a good organization-specific document structure is, therefore, important since it can capture a shared understanding among the organizational actors about how certain business processes should be performed. Current tools that generate document type definitions focus on the underlying technology, emphasizing tags created in a single instance document. The tools, thus, fall short of capturing the shared understanding between organizational actors about how a given document type should be represented. We propose a method for inferring organization-specific document structures using multiple instance documents as inputs. The method consists of heuristics that combine individual document definitions, which may have been compiled using standard algorithms. We propose a number of heuristics utilizing artificial intelligence and natural language processing techniques. As the research progresses, the heuristics will be tested on a suite of test cases representing multiple instance documents for different document types. The complete methodology will be implemented as a research prototype
How to Cite
Sengupta, A., & Purao, S. (1). Transitioning Existing Content: inferring organisation-specific documents. Australasian Journal of Information Systems, 8(1).
Research Articles