DBLPLink 2.0 -- An Entity Linker for the DBLP Scholarly Knowledge Graph

2025-07-30

AI Summary

Linking newly introduced entity types—particularly dblp:Stream (publication venues)—in the DBLP 2025 knowledge graph poses a significant challenge due to the absence of labeled training data and the sparsity of such entities in existing embeddings. Method: We propose a zero-shot entity linking approach that leverages large language models (LLMs) to rerank candidate entities generated by knowledge graph (KG) embedding models. Specifically, we utilize the log-probability of the “yes” token at the penultimate LLM layer to jointly encode RDF semantics and contextual understanding, bypassing the need for supervision. Contribution/Results: The method enables end-to-end disambiguation and alignment of diverse entities—including authors, organizations, and conference streams—within the DBLP RDF schema. Experiments demonstrate substantial accuracy gains in unsupervised settings, especially for dblp:Stream entities, thereby establishing a transferable zero-shot linking paradigm for dynamic expansion of scholarly knowledge graphs.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Description Logics
• Natural Language Processing: Information Extraction

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In this work we present an entity linker for DBLP's 2025 version of RDF-based Knowledge Graph. Compared to the 2022 version, DBLP now considers publication venues as a new entity type called dblp:Stream. In the earlier version of DBLPLink, we trained KG-embeddings and re-rankers on a dataset to produce entity linkings. In contrast, in this work, we develop a zero-shot entity linker using LLMs using a novel method, where we re-rank candidate entities based on the log-probabilities of the "yes" token output at the penultimate layer of the LLM.

Problem

• Develops a zero-shot entity linker for DBLP
• Introduces dblp:Stream as a new entity type
• Reranks candidates using LLM log-probabilities

Innovation

• Zero-shot entity linker using LLMs
• Re-rank candidates via LLM log-probabilities
• DBLP:Stream as new entity type