Towards Event-based Discourse Analysis of Biomedical Text

Annotating biomedical text with discourse-level information is a well-studied topic. Several research efforts have annotated textual zones (e.g., sentences or clauses) with information about rhetorical status, whilst other efforts have linked and classified sets of text spans according to the type of discourse relation holding between them. A relatively new approach has involved annotating meta-knowledge (i.e., rhetorical intent and other types of information concerning interpretation) at the level of bio-events, which are structured representations of pieces of biomedical knowledge. In this paper, we report on the examination and comparison of transitions and patterns of event metaknowledge values that occur in both abstracts and full papers. Our analysis highlights a number of specific characteristics of event-level discourse patterns, as well as several noticeable differences between the types of patterns that occur in abstracts and full papers.