Skip to content
Visualisation of copyright evidence and/or enhance the Copyright Wiki.
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
gradle/wrapper
img
rdfFiles
resultCache
src/main
.gitignore
README.md
build.gradle
gradle.properties
gradlew
gradlew.bat
runServer.bat
runServer.sh
settings.gradle

README.md

©-graph project

This project was developed during the euHackathon 2016, whose objective was to develop tools related to the visualisation of copyright evidence and/or enhance the Copyright Wiki.

There were several sites which served as data sources for developing these visualisations. In addition to the already mentioned Copyright Wiki we also had a collection of studies at the euHackathon resources page.

1. Introduction

We decided to improve the visualisation tools currently available at the wiki by developing two new features:

  • Visualise copyright studies with the ability to group them by several criteria.
  • For each of the resulting nodes of these diagrams, display a tag cloud which contains the most mentioned topics inside the studies that made up these nodes.

When we tought about which kind of grouping criteria would be most useful to copyright policy makers we came up with the following:

  • Grouping by country
  • Grouping by author
  • Grouping by categories

1.1. Interpretation

The groups of studies are displayed trhough graph-style visualisations.

  • The nodes represents the number of studies belonging to a certain group (same contry, same author, etc.)
  • The width and color of the edges between two nodes represents the number of citations from studies in the first group to the second one.

Diagram explanation

1.2. Value

Although it is possible for the wiki right now to display a list of studies belonging to a certain country, author or category, there is no way to visualise the relations created by the studies between these groups.

The kind of visualisations we developed helps to discover relationships between the published copyright studies that otherwise wouldn't be possible.

2. Grouping studies

Visualisation tools right now don't have support for grouping studies.

Previous year

This is the final result of our work:

Current state

2.1. Grouping by country

With this visualisation we can discover the following information:

  • The width of the arrow which connects countries A and B is proportional to the number of studies from A referencing studies from B. This shows us which countries quote which.
  • The tag cloud shows which topics are treated more often in the studies released in this country.

Group by country

2.2. Grouping by author

With this visualisation we can discover the following information:

  • The width of the arrow which connects authors A and B is proportional to the number of studies published by A referencing studies published by B. This helps discover which authors quote which, who collaborates together in studies, etc. This authors may also work in the same topics...

  • The tag cloud shows the topics each author is more interested in...

Group by author

2.3. Grouping by category

(Not yet implemented)

3. Data sets used

To generate the data used in this project we use technologies from the Semantic Web. Specifically, we exported all the pages from the Copyright Wiki to RDF format and imported them into an SPARQL Endpoint. Then, using Apache Jena library on Java we performed semantic queries over the dataset. The use of Semantic Web technologies allowed us to obtain complex results about relations with simple SPARQL queries. An overview of the System Architecture can be seen on the Figure below.

System Architecture The data showed in the group by country and group by author visualisations is real, while the terms showed inside the tag cloud are invented (it's a collection of terms we arbitrarily picked up from the wiki).

4. How to run the project

Requirements:

  • Java
  • NPM (Bundled with Node.JS)
  • Linux, Windows or MacOS.

Steps:

  1. Download Jena-Fuseki server and create a dataset with the files provided in the rdfFiles folder. Leave the server running.
  2. Go to src/main/java/resources/frontend and execute npm install, so the Frontend dependencies are downloaded.
  3. Execute the Web Server by executing runServer.sh or runServer.bat (Linux/Mac or Windows).
  4. Access 'http://localhost:3030' and enjoy :)

Note: Due to having to calculate the data, the first times you use the system it can take long to render the nodes. Next times it will be almost immediate.

You can’t perform that action at this time.