Skip to content
@EnexaProject

EnexaProject

Efficient Explainable Learning on Knowledge Graphs (ENEXA)

ENEXA is a European project focused on developing human-centered explainable machine learning (ML) approaches for real-world knowledge graphs. Funded by the European Union's Horizon Europe program (grant number 101070305), ENEXA aims to bridge the gap between theoretical guarantees and practical application of explainable ML on knowledge graphs.  

ENEXA Project Website
ENEXA Documentation

Project Goals

  • Achieve human-centered, transparent, and explainable AI for ethical development of digital and industrial solutions.
  • Develop scalable, transparent, and explainable hybrid machine learning algorithms by combining symbolic and sub-symbolic learning techniques.
  • Leverage the rich semantics of knowledge graphs as a powerful knowledge representation mechanism across various domains and industries.
  • Address the limitations of existing explainable ML approaches for knowledge graphs, which often struggle with real-world data's scale, incompleteness, and inconsistencies.

Key Innovations

  • Concurrent exploitation of different knowledge graph representations: ENEXA devises novel ML approaches that maintain formal guarantees while concurrently utilizing various knowledge graph representations like formal logics, embeddings, and tensors. This approach aims to achieve significant advancements in the scalability of ML on knowledge graphs.
  • Human-centered co-construction for explainability: ENEXA focuses on developing explainability techniques that promote conversation between humans and machines. This co-construction approach facilitates the creation of human-understandable explanations tailored to user needs.

Project Applications

The project plans to deploy its methods in three crucial European sectors:

  • Business Services  
  • Geospatial Intelligence
  • Brand Marketing

This README provides a high-level overview of the ENEXA project. For detailed technical information, please refer to the individual repository documentation for each module. You can reach the project documentation from this link.

Popular repositories Loading

  1. enexa-tensor-reasoning enexa-tensor-reasoning Public

    Reasoning over knowledge graphs using tensor-based methods

    Python 2 1

  2. enexaproject.github.io enexaproject.github.io Public

    The Website for the ENEXA project

    SCSS 1

  3. enexa-service enexa-service Public

    The core ENEXA service providing functionalities.

    Java

  4. enexa-utils enexa-utils Public

    Utility functions used across various modules

    Shell

  5. enexa-example-module enexa-example-module Public

    Example usage scenarios for various functionalities

    Makefile

  6. enexa-dice-embeddings enexa-dice-embeddings Public

    Techniques for generating knowledge graph embeddings

    Makefile

Repositories

Showing 10 of 18 repositories

Top languages

Loading…

Most used topics

Loading…