Detecting and using document structure in scientific text

The detection of discourse structure of scientific documents is important for a number of tasks, including biocuration efforts, text summarisation, and the creation of improved formats for scientific publishing. Currently, many parallel efforts exist to detect a range of discourse elements at different levels of granularity, and for different purposes, including extraction of information from complex documents, alignment of parallel corpora across languages, and support for document summarization (particularly multi-document summarization). Another interesting class of applications comes from "bibliometrics" and "scientometrics". For example, for analysis of argument structure in full text articles from the scientific literature, it may be important to know where a particular reference is cited or where a particular statement is made (Background, Discussion, etc.). Another application might include tracking over time where (in what sections) an entity or concept is mentioned, to determine whether the mentions migrate from research claims into the "Background" or eventually to the "Methods" sections of articles, as the concept moves from "foreground" (subject of the research) to "background". In this panel we would like to, explore compare, contrast and evaluate different scientific discourse annotation schemes and tools, in order to answer questions such as: