Making DIA Accessible to Non-Experts: Designing a Visual Programming Language for Document Image Analysis

The research field of Document Image Analysis (DIA) produces state-of-the-art methods that are intended to analyze digitized documents, a substantial portion of which lie in the realm of humanities. However, applying the methods, more often than not, requires some programming knowledge. This results in low accessibility and, consequently, many of the methods, if not most, are not used by scholars in the humanities. There is a distinct need to bridge this gap and make DIA methods more accessible to end users. In this context, we discuss the design of a Visual Programming Language (VPL) paradigm for an integrated DIA application environment targeting end-users that are not capable of programming. We elucidate in detail the various design considerations that went into the VPL's development and provide a description of our prototype VPL environment, named, AMAP and, finally, discuss future work.