Embedding Logical Queries on Knowledge Graphs

Learning low-dimensional embeddings of knowledge graphs is a powerful approach used to predict unobserved or missing edges between entities. However, an open challenge in this area is developing techniques that can go beyond simple edge prediction and handle more complex logical queries, which might involve multiple unobserved edges, entities, and variables. For instance, given an incomplete biological knowledge graph, we might want to predict "em what drugs are likely to target proteins involved with both diseases X and Y?" -- a query that requires reasoning about all possible proteins that {\em might} interact with diseases X and Y. Here we introduce a framework to efficiently make predictions about conjunctive logical queries -- a flexible but tractable subset of first-order logic -- on incomplete knowledge graphs. In our approach, we embed graph nodes in a low-dimensional space and represent logical operators as learned geometric operations (e.g., translation, rotation) in this embedding space. By performing logical operations within a low-dimensional embedding space, our approach achieves a time complexity that is linear in the number of query variables, compared to the exponential complexity required by a naive enumeration-based approach. We demonstrate the utility of this framework in two application studies on real-world datasets with millions of relations: predicting logical relationships in a network of drug-gene-disease interactions and in a graph-based representation of social interactions derived from a popular web forum.

PDF Abstract NeurIPS 2018 PDF NeurIPS 2018 Abstract


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Complex Query Answering FB15k GQE MRR 1p 0.546 # 7
MRR 2p 0.153 # 7
MRR 3p 0.108 # 6
MRR 2i 0.397 # 7
MRR 3i 0.514 # 7
MRR pi 0.276 # 6
MRR ip 0.191 # 7
MRR 2u 0.221 # 7
MRR up 0.116 # 6
Complex Query Answering FB15k-237 GQE MRR 1p 0.35 # 6
MRR 2p 0.072 # 6
MRR 3p 0.053 # 6
MRR 2i 0.233 # 6
MRR 3i 0.346 # 7
MRR pi 0.165 # 6
MRR ip 0.107 # 6
MRR 2u 0.082 # 6
MRR up 0.057 # 6


No methods listed for this paper. Add relevant methods here