Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Computation and Language
Title: Editing Language Model-based Knowledge Graph Embeddings
(Submitted on 25 Jan 2023 (v1), last revised 19 Dec 2023 (this version, v8))
Abstract: Recently decades have witnessed the empirical success of framing Knowledge Graph (KG) embeddings via language models. However, language model-based KG embeddings are usually deployed as static artifacts, making them difficult to modify post-deployment without re-training after deployment. To address this issue, we propose a new task of editing language model-based KG embeddings in this paper. This task is designed to facilitate rapid, data-efficient updates to KG embeddings without compromising the performance of other aspects. We build four new datasets: E-FB15k237, A-FB15k237, E-WN18RR, and A-WN18RR, and evaluate several knowledge editing baselines demonstrating the limited ability of previous models to handle the proposed challenging task. We further propose a simple yet strong baseline dubbed KGEditor, which utilizes additional parametric layers of the hypernetwork to edit/add facts. Our comprehensive experimental results reveal that KGEditor excels in updating specific facts without impacting the overall performance, even when faced with limited training resources. Code and datasets are available in this https URL
Submission history
From: Ningyu Zhang [view email][v1] Wed, 25 Jan 2023 04:45:06 GMT (1036kb,D)
[v2] Mon, 6 Feb 2023 09:10:32 GMT (1036kb,D)
[v3] Mon, 13 Feb 2023 13:16:56 GMT (1036kb,D)
[v4] Sat, 25 Mar 2023 12:44:35 GMT (1037kb,D)
[v5] Wed, 16 Aug 2023 10:57:58 GMT (9958kb,D)
[v6] Wed, 15 Nov 2023 14:00:17 GMT (9987kb,D)
[v7] Mon, 18 Dec 2023 16:09:49 GMT (9838kb,D)
[v8] Tue, 19 Dec 2023 13:21:28 GMT (9838kb,D)
Link back to: arXiv, form interface, contact.