The Open University
7 files

Materials for mapping study: structured data to RDF

Version 3 2023-05-30, 13:52
Version 2 2023-05-22, 10:25
Version 1 2022-11-07, 12:09
posted on 2023-05-30, 13:52 authored by Paul WarrenPaul Warren

These files were used with participants in a study of the experience of using YARRRML and SPARQL Anything to map structured data, e.g. CSV, JSON and XML, to RDF.  YARRRML is a mapping language, whereas SPARQL Anything is an extension of SPARQL which uses the CONSTRUCT query to define the mapping.  

The study was a between-participants study, i.e. some participants were asked to use YARRRML and other participants were asked to use SPARQL Anything.  Participants were observed undertaking a number of mapping tasks; the same mapping tasks were used for both conditions. Participants took part remotely, using Microsoft Teams, and their screen activity was recorded, along with any comments they made.  Subsequent analysis of the recordings enabled a qualitative study of their experiences.

The files are::

* YARRRML tutorial

* Information for YARRRML study participants, including questions

* SPARQL Anything tutorial

* Information for SPARQL Anything study participants, including questions



Please note that the two tutorials are not comprehensive, but contain only the information required by participants in the study.

The two zip files were as sent out  to participants immediately prior to the study.  They contain:

* The data files, i.e. artist_data.csv, artist38.csv; artwork.json, artwork.xml, artworkAttributes.xml.

* For the YARRRML participants, 1.yml, 2.yml etc, which contain the YARRRML code requiring completion; and similarly for the SPARQL Anything participants, 1.sparql, 2.sparql etc, which contain the SPARQL Anything code requiring completion.

* 1.ont, 2.ont etc., which contain the required output for each question.

* A 'bak' folder which contain either 1.yml, 2.yml etc., or 1.sparql, 2.sparql etc., as appropriate.  These were provided in case a participant wished to start a question again with a 'clean slate'.  

The .yml, .sparql and .ont files in the two zip files are numbered as in the paper.  Some participants were sent alternative zip files with the order of questions 3, 4, 5 and the order of questions 6, 7, 8 permuted.


Usage metrics

    Faculty of Science, Technology, Engineering and Mathematics (STEM)



    Ref. manager