Neuro-Symbolic AI Tutorial for Industry 4.0

Welcome to the NeSyAI4 Tutorial for Industry 4.0 Repository at the International Semantic Web Conference!

This repository serves as a resource for attendees of our hands-on tutorial on Neuro-Symbolic AI for Industry 4.0. In this tutorial, we explore the fusion of semantic web technology and machine learning, demonstrating practical applications in the context of Industry 4.0.

During the hands-on session at the International Semantic Web Conference, participants will gain practical experience in applying Neuro-Symbolic AI to Industry 4.0. They'll explore semantic technology, ontologies, and deep learning to address real-world manufacturing challenges, using Bosch scenarios as examples.

Organizers

This tutorial is organized by Robert Bosch researhers in the field of Neuro-Symbolic AI and Industry 4.0:

• Diego Rincon-Yanez
  Affiliation: Bosch Center for AI Email: fixed-term.diego.rincon [at] de.bosch.com, rinconyanezd+iswc [at] gmail.com

• Irlan Grangel-Gonzalez
  Affiliation: Bosch Corporate Research Email: irlan.grangelgonzalez [at] de.bosch.com

• Mohamed H. Gad-elrab
  Affiliation: Bosch Corporate Research Email: Mohamed.Gad-Elrab [at] de.bosch.com

• Yuqicheng Zhu Affiliation: Bosch Center for AI Email: Yuqicheng.Zhu [at] de.bosch.com

• Evgeny Kharlamov
  Affiliation: Bosch Corporate Research Email: Evgeny.Kharlamov [at] de.bosch.com

Feel free to reach out to us with any questions or concerns regarding the tutorial. We look forward to your participation!

Requirements

Make sure you have the following installed software before starting the hands-on session:

• Git Client
• Text Editor
• Docker Engine Instance

Running the example using docker compose

Download the prepared compose.yaml file from the repository via wget and execute the command using the utility.

git clone https://github.com/d1egoprog/ISWC23-I40Tutorial.git
docker compose -p iswc23-i40 up -d

The previous lines will download the repository (specially for the example files) and run the following docker compose file automatically in a pipeline fashion.

version: '3.7'

services:
  mapping:
    build: 
      context: ./1.${COMPONENT_1}/.
      args:
        - VERSION=${VERSION}
        - COMPONENT_NAME=${COMPONENT_1}
    environment:
      - ONTOLOGY=${ONTOLOGY}
      - OUTPUT_KG=${OUTPUT_KG}
      - INPUT_MAPPING=${INPUT_MAPPING}
    volumes:
      - ./files/datasources:/opt/${COMPONENT_1}/datasources
      - ./files/ontologies:/opt/${COMPONENT_1}/ontologies
      - ./files/mappings:/opt/${COMPONENT_1}/mappings
      - ./files/output:/opt/${COMPONENT_1}/output
    image: ${PROJECT_PREFIX}-${COMPONENT_1}:${VERSION}
  querying:
    environment:
      - INPUT_KG=${OUTPUT_KG}_merged.ttl
    volumes:
    - ./files/output:/opt/jena-fuseki/output
    - ./files/logs:/opt/jena-fuseki/logs
    ports:
      - 3030:3030
    image: d1egoprog/jena-fuseki:4.9.0-web
    depends_on:
      mapping:
        condition: service_completed_successfully
  visualization:
    image: d1egoprog/webvowl:1.1.7
    ports:
        - 8080:8080
    restart: always
    depends_on:
      mapping:
        condition: service_completed_successfully

Running on Mac M-type chips

For running on M1 and M2 the parameter platform: linux/amd64 needs to be added to the service in the docker compose file at the same level from the build statement.

services:
  mapping:
    platform: linux/amd64
    build: 
      context: ./1.${COMPONENT_1}/.
      args:
        - VERSION=${VERSION}
        - COMPONENT_NAME=${COMPONENT_1}

Feedback and Contributions:

We welcome your feedback and contributions to this repository. If you have suggestions, improvements, or additional resources to share, please feel free to open issues, submit pull requests, or reach out to the tutorial organizers.

Enjoy your learning journey with Neuro-Symbolic AI for Industry 4.0!