Discriminative Gaifman Models
This work tackles the problem of making relational machine learning more tractable and robust for knowledge base applications, though it appears incremental by building on existing relational approaches.
The paper introduces discriminative Gaifman models, a novel family of relational machine learning models that learn feature representations from local, bounded-size neighborhoods in knowledge bases to address tractability and overfitting issues, and applies them to large-scale relational learning problems.
We present discriminative Gaifman models, a novel family of relational machine learning models. Gaifman models learn feature representations bottom up from representations of locally connected and bounded-size regions of knowledge bases (KBs). Considering local and bounded-size neighborhoods of knowledge bases renders logical inference and learning tractable, mitigates the problem of overfitting, and facilitates weight sharing. Gaifman models sample neighborhoods of knowledge bases so as to make the learned relational models more robust to missing objects and relations which is a common situation in open-world KBs. We present the core ideas of Gaifman models and apply them to large-scale relational learning problems. We also discuss the ways in which Gaifman models relate to some existing relational machine learning approaches.