Alfalab research and development teams, comprised of humanities researchers and IT engineers, currently work on developing two demonstrators— TextLab and SpaceLab.
With these demonstrators, Alfalab addresses the problem of sustainability and/or applicability of tools tailored to heterogeneous data, and posits that tools should be defined as task-oriented service components. A distinctive advantage of such an approach is that it takes the particular conditions of humanities as a starting point, rather than trying to impose models from tools developed in other scientific domains.
TextLab includes a number of digital tools for textual exploration, such as a named entity recognition service for onomastic studies; a service for transcription and annotation of textual sources, and; an inductive support service for auto-suggesting textual annotation. The later tool draws on successful practices of social networking sites, such as LinkedIn and Facebook, and it supports researchers in transcribing and/or annotating texts by providing suggestions in the following form:
Other researchers annotating the words [....] added the following annotation [....]
Suggestions are based on word form distance computation and context matching. This tool should thus show the potential of task driven services to support digital scholarship in contexts of heterogeneous and ambiguous data. The heterogeneity in this case is ’semi-solved’ through word distance and context computation; the problem of ambiguous data is ’semi-solved’ by the human supervised annotation suggesting algorithm.
Similar to TextLab, SpaceLab will also demonstrate some of the principles of coping with heterogeneous and ambiguous data.
SpaceLab offers tools for (and tutorials on) applying Geographical Information Systems (GIS) to historical materials. An example of this is a web-based map annotation tool currently used in two research projects. Within the digital rural microtoponyms project of the Meertens Institute, the tool is used to align (‘georeference’) maps that were annotated by hand. Within the MAPS project run at the University of Utrecht and the National Archives (external partners of the KNAW and Alfalab), digitized maps are employed in a wide variety of cartography methods and styles (for instance, the corpus in use incorporates 17th century townscapes).
Another application of the image-annotation tool is being developed in connection to TextLab, where this tool can be used to annotate the visual landscape represented by manuscript facsimile (for instance, referencing and identifying illuminations in the manuscripts).