Semanlink - Survey / Review AND Domain Knowledge + Deep Learning

Survey / Review AND Domain Knowledge + Deep Learning

Common descendants

4 Documents

Graph ML in 2023: The State of Affairs | by Michael Galkin | Towards Data Science

Tags:

2023-01-03 About

A review of some techniques for inclusion of domain‐knowledge into deep neural networks

Tags:

2022-05-19 About

The knowledge graph as the default data model for learning on heterogeneous knowledge (2017)

Tags:

2019-08-22 About

Knowledge Graph Embedding: A Survey of Approaches and Applications - IEEE (2017)

Tags:

- "facts only" based embeddings
    - Translational Distance Models, distance-based scoring functions
       - TransE and co: entities and relations as points in vector space
           - TransH: relation-specific hyperplanes.
       - Gaussian Embeddings: entities and relations as random variables
    - Semantic matching models, similarity-based scoring functions
        - RESCAL (entities as a vector, relation as a matrix) and familiy
            - DistMult (relation as diagonal matrix)
            - Holographic Embeddings (HolE)
            - etc.
        - Matching with Neural Networks
            - SME, etc
- incorporating other info: entity types, relation paths,
textual descriptions, as well as logical rules
    - entity types
        - semantically smooth embedding
(SSE):  entities of the same type to stay close
to each other in the embedding space
        - type-embodied knowledge representation
learning (TKRL), which can handle hierarchical
entity categories and multiple category labels
- Applications in downstream tasks

2019-05-27 About