Relation Adversarial Network for Low Resource Knowledge Graph Completion

Abstract

Knowledge Graph Completion (KGC) has been proposed to improve Knowledge Graphs by filling in missing connections via link prediction or relation extraction. One of the main difficulties for KGC is a low resource problem. Previous approaches assume sufficient training triples to learn versatile vectors for entities and relations, or a satisfactory number of labeled sentences to train a competent relation extraction model. However, low resource relations are very common in KGs, and those newly added relations often do not have many known samples for training. In this work, we aim at predicting new facts under a challenging setting where only limited training instances are available. We propose a general framework called Weighted Relation Adversarial Network, which utilizes an adversarial procedure to help adapt knowledge/features learned from high resource relations to different but related low resource relations. Specifically, the framework takes advantage of a relation discriminator to distinguish between samples from different relations, and help learn relation-invariant features more transferable from source relations to target relations. Experimental results show that the proposed approach outperforms previous methods regarding low resource settings for both link prediction and relation extraction.

Reasoning Through Memorization: Nearest Neighbor Knowledge Graph Embeddings

“…), previous studies utilize KG embeddings or textual encoding to represent triple and leverage a pre-defined scoring function to those vectors. With pretrained encoders, those approaches such as KG-BERT [Yao et al, 2019] are generalizable enough and robust to the incompleteness; however, they have to costly scoring of all possible triples in inference. In this paper, we simply leverage masked entity modeling for link prediction, which makes the model predict the correct entity e k like the Masked Language Model (MLM) task.…”

“…Knowledge Graphs (KGs) organize facts in a structured way as triples in the form of <subject, predicate, object>, abridged as (s, p, o), where s and o denote entities and p builds relations between entities. Most KGs are far from complete due to the emerging entities and their relations in real-world applications; hence KG completion-the problem of extending a KG with missing triples-has appeal to researchers [Lin et al, 2015;Zhang et al, 2019a;Zhang et al, 2020;Zhang et al, 2021a;Qi et al, 2021;Zhang et al, 2021b].…”

ZJUKLAB at SemEval-2021 Task 4: Negative Augmentation with Language Model for Reading Comprehension of Abstract Meaning

“…• How to help models? We could use our approach of NAL to increase the weight of the knowledge learned in the pre-training task or leverage external knowledge (Zhang et al, 2019(Zhang et al, , 2020bYu et al, 2020;Zhang et al, 2020a).…”

OpenUE: An Open Toolkit of Universal Extraction from Text

“…Relational Triple Extraction is an essential task in Information Extraction (IE) for Natural Language Processing (NLP) and Knowledge Graph (KG) (Zhang et al, 2018b;Yu et al, 2017;Nan et al, 2020;Zhang et al, 2020a;Ye et al, 2020;Zhang et al, 2020b), which is aimed at detecting a pair of entities along with their relations from unstructured text. For instance, there is a sentence "Paris is known as the romantic capital of France.…”