TDG DanielAyala Committed to research! TDG Site Manager 1.1

ISWC 2017

In this page we have made available the following resources:

  • The source code of the latest version of our proposal can be found in the following repository.
  • A jar with the latest version of our proposal can be found here, ready to be applied to our datasets, or any other dataset, including a .bat file with the command used to start the tests.
  • The datasets used for learning and labelling can be found here. The zip file contains datasets from the 10 sources we used, separated into 10 folds. Each fold contains several json files, corresponding to a labelled structure with information.
  • The models we created from the learning sets can be found here. The zip file contains several folders corresponding to the model applied to each combination of two folds. Each model contains the classifiers used by the ensemble technique, both instance and literal ones, in their hint-based and hint-free versions.
  • The detailed results of our experimentation can be found here. The zip file contains 45 folders and a csv file. The 45folders correspond to the labelling of each combination of two folds. Each folder contains a list of files. Each contains a structure with the true and the infered labels. The csv file contains the measures of each experiment and class.