Distributional Negative Sampling for Knowledge Base Completion
This addresses a training inefficiency for researchers and practitioners in knowledge base completion, though it is incremental as it builds on existing methods.
The paper tackles the problem of generating nonsensical negative examples in Knowledge Base Completion by proposing Distributional Negative Sampling, which creates meaningful negatives, resulting in significant improvements in Mean Reciprocal Rank values across benchmarks.
State-of-the-art approaches for Knowledge Base Completion (KBC) exploit deep neural networks trained with both false and true assertions: positive assertions are explicitly taken from the knowledge base, whereas negative ones are generated by random sampling of entities. In this paper, we argue that random sampling is not a good training strategy since it is highly likely to generate a huge number of nonsensical assertions during training, which does not provide relevant training signal to the system. Hence, it slows down the learning process and decreases accuracy. To address this issue, we propose an alternative approach called Distributional Negative Sampling that generates meaningful negative examples which are highly likely to be false. Our approach achieves a significant improvement in Mean Reciprocal Rank values amongst two different KBC algorithms in three standard academic benchmarks.