NLP_ThumbnailAnnotator

M.Sc. Project in Language Technology at Uni Hamburg from Florian Schneider

---

Project maintained by uhh-lt Hosted on GitHub Pages — Theme by mattgraham

Admin Guide

Please note that an internet connection is crucial mandatory! Also be aware that this application (specially for testing) needs a lot of heap space (~3-4 GB) due to the models used in the UIMA Pipeline.

Install prerequisites

• Java 8
• Maven 3
• redis-server >= 3.2
• npm
• nginx
• docker
• docker-compose >= 3.0

sudo apt update && sudo apt upgrade -y 
sudo apt install -y oracle-java8-installer oracle-java8-set-default maven npm nginx redis-server docker docker-compose

Clone repository

  git clone https://github.com/floschne/NLP_ThumbnailAnnotator.git && cd NLP_ThumbnailAnnotator

Apply JWNL Properties File ‘Bugfix’

When using WordNet with DKPro WSD, a hard coded, non-relative path to the WordNet files has to be set in a properties file. Since the absolute location of those WordNet files is not known at runtime, the temporary solution is to copy the files to /tmp/.

  cp -r thumbnailAnnotator.parent/thumbnailAnnotator.core/src/main/resources/WordNet-3.0 /tmp/

Run with docker-compose

• Package and the java-based REST API

  cd thumbnailAnnotator.parent
  mvn clean package spring-boot:repackage
  cd ..
  

• Run

  docker-compose up --build
  

• Open Browser

  • localhost:80 for Web-Frontend
  • localhost:8081 for Swagger-UI

Run on a local (Debian based) machine (without Docker)

• Start the redis-server

  sudo service redis-server start
  

• Package and run the java-based REST API

  cd thumbnailAnnotator.parent
  mvn clean package spring-boot:repackage
  java -jar -Dspring.profiles.active=local thumbnailAnnotator.api/target/thumbnailAnnotator.api-0.1.BETA.jar &
  cd ..
  

• Build and run the Vue.js-based Web-Frontend

  cd app
  npm run build
  sudo cp -r dist index.html /var/www/html/
  sudo service nginx restart
  cd ..
  

• Open Browser

  • localhost:80 for Web-Frontend
  • localhost:8081 for Swagger-UI