InteractE: Improving Convolution-based Knowledge Graph Embeddings by Increasing Feature Interactions
By Shikhar Vashishth, Soumya Sanyal, Vikram Nitin, Nilesh Agrawal, Partha P. Talukdar. AAAI Conference on Artificial Intelligence, 2019
399 citations recorded by the source corpus.
Most existing knowledge graphs suffer from incompleteness, which can be alleviated by inferring missing links based on known facts. One popular way to accomplish this is to generate low-dimensional embeddings of entities and relations, and use these to make inferences. ConvE, a recently proposed approach, applies convolutional filters on 2D reshapings of entity and relation embeddings in order to capture rich interactions between their components. However, the number of interactions that ConvE can capture is limited. In this paper, we analyze how increasing the number of these interactions affects link prediction performance, and utilize our observations to propose InteractE. InteractE is based on three key ideas -- feature permutation, a novel feature reshaping, and circular convolution. Through extensive experiments, we find that InteractE outperforms state-of-the-art convolutional link prediction baselines on FB15k-237. Further, InteractE achieves an MRR score that is 9%, 7.5%, and 23% better than ConvE on the FB15k-237, WN18RR and YAGO3-10 datasets respectively. The results validate our central hypothesis -- that increasing feature interaction is beneficial to link prediction performance. We make the source code of InteractE available to encourage reproducible research.
Subfields: Artificial Intelligence, Management Science and Operations Research
View Paper: https://www.semanticscholar.org/paper/158d9ced2892e484e7d62b660ef2bb81c4bdfc55
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Soumen Chakrabarti
- Partha P. Talukdar
- Apoorv Saxena
- Denilson Barbosa
- Xirui Ke
- Haipeng Ding
- Lingxi Zhang
- Bo Chen
In Management Science and Operations Research
Citations in the corpus
- Improving Multi-hop Question Answering over Knowledge Graphs using Knowledge Base Embeddings (2020, background): “InteractE (Vashishth et al., 2019) improves upon ConvE by increasing feature interaction.”
- Knowledge Graph Embedding for Link Prediction (2020, methodology): “In both experiments, we report results for the three best-performing models of each family, that is, ComplEx [67] for bilinear family, HAKE [81] for the geometric family, and [69] for the Deep Learning family; we use the same hyperparameters and number of training epochs in each training process.”
- Mind the Pad - CNNs can Develop Blind Spots (2020, background): “Problem-specific constraints can dictate different choices [30; 31; 36].”
- Question Answering Over Temporal Knowledge Graphs (2021, methodology): “Several methods have been proposed in the literature to embed KGs (Bordes et al. 2013, Trouillon et al. 2016, Vashishth et al. 2020).”
- Neural, symbolic and neural-symbolic reasoning on knowledge graphs (2021, methodology): “ConvR (Jiang et al., 2019) extends the global fi lters in ConvE to relation-speci fi c fi lters, and InteractE (Vashishth et al., 2020a) captures more interactions by additional convolution operations.”