An official website of the United States government
Knowledge graph embedding (KGE) is a technique for learning continuous embeddings for entities and relations in the knowledge graph. Due to its benefit to a variety of downstream tasks such as knowledge graph completion, question answering and recommendation, KGE has gained significant attention recently. Despite its effectiveness in a benign environment, KGE's robustness to adversarial attacks is not well-studied. Existing attack methods on graph data cannot be directly applied to attack the embeddings of knowledge graph due to its heterogeneity. To fill this gap, we propose a collection of data poisoning attack strategies, which can effectively manipulate the plausibility of arbitrary targeted facts in a knowledge graph by adding or deleting facts on the graph. The effectiveness and efficiency of the proposed attack strategies are verified by extensive evaluations on two widely-used benchmarks.
more »
« less
This content will become publicly available on January 1, 2026
Predicting Interactions in the Weapons of Mass Destruction Knowledge Graphs. In this paper, we apply graph machine learning methods to predict unseen interactions within the Weapons of Mass Destruction (WMD) dataset, developed by DARPA and IARPA. This dataset captures complex online activities, including sales, purchases, and forum discussions, with a focus on topics such as weapons, explosives, and other sensitive subjects. We represent the data as a knowledge graph, where nodes correspond to entities and edges denote relationships between them. Among various knowledge graph embedding techniques and graph neural networks, semantic matching models like DistMult demonstrate the ability to accurately predict 84% of relations, particularly due to their strength in capturing the one-to-many relationships common in the WMD data. To streamline the analysis, we implement an automated pipeline that stores the knowledge graph in a Neo4j database, extracts subgraphs using Cypher queries, trains knowledge graph embedding models on these subgraphs, predicts links, and reintegrates high-confidence edges back into the main graph.
- Award ID(s):
- 2528805
- PAR ID:
- 10614656
- Publisher / Repository:
- Springer Nature Switzerland
- Date Published:
- Page Range / eLocation ID:
- 211 to 223
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
