This repository contains all material that was used for the user-study described in the ESWC 2016 paper "Comparing Vocabulary Term Recommendations using Association Rules and Learning To Rank: A User Study".
Contents:
-
dataForTasks/: This folder contains the data that was used for the three modeling tasks including the practice data (museumData, musicData, weapons, publicationData), as well as the data's initial R2RML models (in Turtle syntax) that generate the Karma graph for the data. In addition, it comprises two sub-folders: one contains the gold standard RDF representation for each data set that was conducted by five different LOD experts and the other one contains the participants' LOD representations (the subfolder number specifies the participants' arbitrary ID)
-
taskDocs/: This folder contains the instruction documents explaining the assignments the participants had to complete in each modeling task. It also comprises a Karma-How-To Guide to help participants if they were not familiar with the Karma operations needed to complete the modeling tasks.
-
usedVocabularies/: This folder contains all vocabularies that were used to model the data for all modeling tasks
-
satisfactionQuestionnaire.pdf: This PDF comprises the questionnaire that was presented to the participants to indicate their level of satisfaction regarding the recommendation approaches Learning To Rank and Association Rule Mining
The raw results of the user-study can be found under the following DOI: http://dx.doi.org/10.7802/1206