| Abstract |
My presentation will be about the CHOICE project (http://www.nwo.nl/CATCH/CHOICE) and will focus
on different modalities for the evaluation of the automatic keywords
suggestions: in the project, we extract keywords from textual documents
describing TV programs with the aim of presenting them to the catalogers who
describe manually the TV programs, as keywords suggestion. We tested different
evaluation paradigms for getting measures about the value of such automatic
annotation propositions. |