======================================================================== A Benchmark For Open Relation Extraction Version 1.0: October 2013 Authors: Filipe Mesquita Jordan Schmidek Denilson Barbosa ======================================================================== (1) CONTENT OF THIS DISTRIBUTION The directory 'data' is organized as follows: binary/ -- Data for the experiment with binary relations. manual/ -- Ground truth was manually annotated. dev/ -- Development sentences. test/ -- Test sentences. automatic/ -- Ground truth was automatically annotated. dev/ -- Development sentences. test/ -- Test sentences. nary/ -- Data for the experiment with n-ary relations. Ground truth was manually annotated. Test sentences only. (2) LIST OF CHANGES (3) EVALUATION FILES & SCRIPTS For instructions on how to run a specific experiment, please refer to the README file for this experiment in the directory 'docs'. Each file explains the format of input and output files as well as how to run the evaluation scripts.