Exploring Web-Based Methods for Syntactic Annotation

In recent years, OpenText.org and BiblicalHumanities.org have, together with other interested parties, drafted an inline XML format to facilitate the open exchange of linguistic annotations of ancient texts. Previously, we presented our rewritten OpenText.org display tools, which work with the new XML format and which enable end-users to visualize syntactic annotations using nested box diagrams. In this paper, we move one step further, exploring the potential usefulness of different display conventions and input methods for the task of generating new syntactic annotations. We focus on three specific areas: (1) the distribution of labour between automatic and manual annotation and its impact on the design of manual annotation tools; (2) the efficiency of different input methods for the two key tasks involved in generating the new XML (i.e. delimiting text spans and assigning attributes) and the merits/demerits of accommodating the process to one primary method; and (3) the relative usefulness of different display conventions in the design of a GUI for a web-based annotator/editor. After surveying existing tools, we will present our own concepts for a new annotation tool. (And if possible, we will allow participants to experiment with different prototypes in order to experience the strengths and weaknesses of the some of the options explored in the paper.)