Hi, Am 23.10.2020 um 23:56 schrieb Richard Eckart de Castilho: > An argument which supports Peter's symmetric approach: > > With the asymetric approach where the begin offset is included in the > interval, but not the end offset, a zero-width annotation at the end of a > document could never be selected via a select.coveredBy which spans the > entire document. > > With the symmetric approach, the zero-width annotation at the end of the > document would be selectable in this case. > > @Peter: using more naive predicate conditions also induces changes on the > non-zero-width version of the predicates. > E.g. that LEFT_ADJACENT implies LEFT_OVERLAPPING and RIGHT_ADJACENT implies > RIGHT_OVERLAPPING. So cases like > [0-1] [1-2] are suddenly OVERLAPPING which I think doesn't make too much > sense... does it? E.g. when I want to > define a rule that Tokens in a document must not overlap, then two Tokens at > [0-1] [1-2] should IMHO be perfectly > valid - i.e. they should not be considered as overlapping... right?
Yes, there needs to be some additional changes then. In my opinion, it's ok if ADJACENT doe not imply OVERLAPPING. Just by listening to it, it sounds also somwhat reasonable. Afterall, it's just a name for a predicate configuration. For zero-length annotations, OVERLAPPING would still apply due to the inner minimal span that is not ADJACENT. Best, Peter > > -- Richard -- Dr. Peter Klügl Head of Text Mining/Machine Learning Averbis GmbH Salzstr. 15 79098 Freiburg Germany Fon: +49 761 708 394 0 Fax: +49 761 708 394 10 Email: [email protected] Web: https://averbis.com Headquarters: Freiburg im Breisgau Register Court: Amtsgericht Freiburg im Breisgau, HRB 701080 Managing Directors: Dr. med. Philipp Daumke, Dr. Kornél Markó
