A First Experiment on Including Text Literals in KGloVe
This is an incremental improvement for knowledge graph embedding research, addressing the limitation of ignoring literal properties.
The paper tackled the problem of incorporating textual literals into knowledge graph embeddings by combining global graph structure with textual information, but the initial experiment did not clearly outperform earlier methods on machine learning tasks.
Graph embedding models produce embedding vectors for entities and relations in Knowledge Graphs, often without taking literal properties into account. We show an initial idea based on the combination of global graph structure with additional information provided by textual information in properties. Our initial experiment shows that this approach might be useful, but does not clearly outperform earlier approaches when evaluated on machine learning tasks.