/cvServer

Server side setup of the LWMap Server like it was used on http://lwmap.uni-koeln.de

Primary LanguageJavaScriptMIT LicenseMIT

cvServer

Context Visualisation Server

Server side setup of the LWMap Server like it was used on http://lwmap.uni-koeln.de/

Its was created as Part of my dissertation: http://kups.ub.uni-koeln.de/id/eprint/7533

Create your own Wikipedia Map:

choose starting-articles and link-language, browse the visualization, filter the interesting Nodes, download and share your visualization.

About

The context visualisations server was created as part of my dissertation.

It originally and hopefully still runs on http://lwmap.uni-koeln.de/ with data form DBpedia 3.9.

The following installation description describes how to setup and Install the system on a Linux (Ubuntu 16.04 LTS should work).

Overview

WorkScheme

Folders

./createData Contains the create Data script, which Build up the base Data for the Application.

./doc contains Documentation

./Java_exec contains the java Routines used to create the nets.

./run contains Nodejs Server scripts

./Settings contains Setting files and Example for Settings see Readme File

./www_content contains the Data For the Web-server

./fuseki Install Fuseki here (install.sh)

./cvManagement Holds Start/restart/Checklog scripts see readme for Configuration

./UserTracking contains a description of the user tracking system an the prototype Database

Dependencies

Java 7, openJDK to Execute Fuseki and the Core Executeable.

nodejs, v0.10 for the Server Programm and the user Tracking

A Webserver(apache, nginx) for the static Data.

Docker

The There is a Docker Testbed including all dependencies.

Execute Image

sudo docker build . -t my:cvServer

sudo docker run --name cvServer -v TriplesInTDB:/srv/cvserver_data/triples/ -p 8090:8090 -p 3030:3030 -p 8080:80 my:cvServer

Replace the TriplesInTDB with the Folder where you executed createData script

Used Ports

There are Three Ports Used:

  • 8080 for the Webfrontend this you can call with your browser
  • 8090 This is the port of the Internal nodeJS Server
  • 3030 Fuseki (RDF Triplestore)

Make shure these ports are unused when starting the Container! For Changes Setting/dockerSettings/local.json must be altered!!

Futher Access

User Tracking

The User Tracking Database can be Access by Opening the MySQl/MariaDB port (usually 3306).

Add: -p:3306

Then you can access the Database with the Database Management tool of you choice.

Extra Mountpoints

There are Folder containing the Data which maybe should not stay only inside the container. you can mount some Folders to make the Data Permanent:

/usr/local/apache2/htdocs/pubfiles

Prepare data

In folder createData is a script that will prepare all the Data form the DBpedia dump 2016.

WARNING Executing the Script will take a lot of diskspace, the Process will consume a lot of RAM (6GB preset) and it will take a lot of time(more than 4 Hours depending on your system).

Download dbpedia Page Links, classes and redirects (for lookups) form dbpedia.

In the project the Internationalized links and the classes Files where Taken from DBpedia.

http://wiki.dbpedia.org/Downloads

Afterwards compress them into TDB-Databases.

http://jena.apache.org/

These files must be linked in the setting/local.ttl

attribute: tdb:location.

The location.ttl can be modified from the example file Settings/fuseki-test.ttl

In the example the TDB stores are in /var/opt/cvserver_data/ , replace it with the Folders of the created TDB stores.

The Triple-store can be on another server (Load balancing).

Manual Installation

configure setting/local.json see setting/readme.txt for instructions.

execute install.sh This will download and unpack Apache fuseki to ./fuseki and SigmaJS, Bootstrap in the www_content Folder, it will also change settings in the settings of www_content so run before moving the www_content.

If you can Proceed take a look at the content of the Scripts and the Dockerfile to get a deeper insight into what happens at installation time!

Nodejs packages

Install npm Packages:

sparql, mysql

Configure Web-Server

Copy the Web-Frontend

example:

cp -r /opt/CVServer/www_content/ /var/www/ContextVis/

The project needs a subfolder pubfiles in the Web Documents where the created gexf files are stored setting/local.json (see setting/readme.txt for instructions)

User tracking:

  • Get an Mysql/MariaDB Database (install it on the machine for example)
  • create Database
  • read Database structure from UserTracking/UserTracking.sql into Database
  • Add Database configuraion to setting/local.json see setting/readme.txt for instructions

Starting The Server

Run with CV Management:

Configure the CV-management see cvManagement/Readme.txt

cvManagement/node_start.sh Start the Node JS Server

cvManagement/node_restart.sh Restarts The nodejs Server if its allready Running

cvManagement/node_checklog.sh tails the log of the Nodejs Server

cvManagement/fuseki_start.sh Start fuseki

cvManagement/fuseki_restart.sh Restarts fukseki

cvManagement/fuseki_checklog.sh tails the log of fuseki