Building Memory with Concept Learning Capabilities from Large-scale Knowledge Base
This work addresses the challenge of handling unseen entities in knowledge bases for AI systems, representing an incremental improvement over previous neural embedding models.
The paper tackles the problem of neural knowledge base embeddings by introducing a framework that models symbolic knowledge and its learning process, achieving superior reasoning performance and enabling learning embeddings for unseen entities from natural language descriptions.
We present a new perspective on neural knowledge base (KB) embeddings, from which we build a framework that can model symbolic knowledge in the KB together with its learning process. We show that this framework well regularizes previous neural KB embedding model for superior performance in reasoning tasks, while having the capabilities of dealing with unseen entities, that is, to learn their embeddings from natural language descriptions, which is very like human's behavior of learning semantic concepts.