Inductive Entity Representations from Text via Link Prediction

Daniel Daza, Michael Cochez, Paul Groth

Research output: Contribution to journalArticlepeer-review


We present a method for learning representations of entities, that uses a Transformer-based architecture as an entity encoder, and link prediction training on a knowledge graph with textual entity descriptions. We demonstrate that our approach can be applied effectively for link prediction in different inductive settings involving entities not seen during training, outperforming related state-of-the-art methods (22% MRR improvement on average). We provide evidence that the learned representations transfer to other tasks that do not require fine-tuning the entity encoder. In an entity classification task we obtain an average improvement of 16% accuracy compared with baselines that also employ pre-trained models. For an information retrieval task, significant improvements of up to 8.8% in NDCG@10 were obtained for natural language queries.
Original languageAmerican English
StatePublished - 2021


Dive into the research topics of 'Inductive Entity Representations from Text via Link Prediction'. Together they form a unique fingerprint.
  • DL: ICAI Discovery Lab

    van Harmelen, F. (CoPI), De Rijke, M. (CoI), Siebert, M. (CoI), Hoekstra, R. (CoPI), Tsatsaronis, G. (CoPI), Groth, P. (CoPI), Cochez, M. (CoI), Pernisch, R. (CoI), Alivanistos, D. (CoI), Mansoury, M. (CoI), van Hoof, H. (CoI), Pal, V. (CoI), Pijnenburg, T. (CoI), Mitra, P. (CoI), Bey, T. (CoI) & de Waard, A. (CoPI)


    Project: Research

Cite this