Knowledge representation learning (KRL) aims to obtain the embedding of entities and relations from the information of knowledge graph (KG). Most existing methods can only model the entities in the training data, while failing to generalize to out-of-knowledge-base (OOKB) entities which only appear in the testing. To solve this issue, one common approach is to train an aggregator by leveraging the auxiliary knowledge such as neighbor information and entity descriptions. In this work, we propose a novel aggregation model called neighborhood transformer (Neighbor-T) to enhance the representations of OOKB entities. Compared with previous methods, Neighbor-T shows effectiveness on neighbor information aggregation because of self-attention mechanism. Experiments demonstrate that our enhanced representation outperforms the state-of-the-art on two knowledge graph completion tasks under OOKB setting: triple classification and entity prediction.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.