Knowledge Fusion via Embeddings from Text, Knowledge Graphs, and Images
arXiv:1704.06084v18 citations
Originality Synthesis-oriented
AI Analysis
This work addresses the challenge of integrating knowledge from multiple modalities for AI systems, but it is incremental as it builds on existing embedding approaches.
The paper tackles the problem of cross-modal knowledge fusion by evaluating different fusion methods on existing embeddings to create unified concept representations across text, knowledge graphs, and images, but does not report specific numerical results.
We present a baseline approach for cross-modal knowledge fusion. Different basic fusion methods are evaluated on existing embedding approaches to show the potential of joining knowledge about certain concepts across modalities in a fused concept representation.