Search papers, labs, and topics across Lattice.
To resolve pervasive entity fragmentation in scientometrics, the authors fine-tuned open-weight embedding models across the Sentence Transformer, Gemma, and Qwen3 families using multi-task contrastive learning on an aligned corpus of ROR, Web of Science, and Crossref funder data. The resulting models achieve over 0.90 accuracy in mapping variant funder names to canonical ROR identifiers, outperforming frontier LLMs like GPT-5.2 and Claude-Sonnet-4.6 by more than 10 percentage points while handling unindexed entities via graph-based similarity clustering. This provides an efficient, scalable blueprint for resolving multilingual real-world entities without relying on costly closed-source API calls.
Open-weight embedding models fine-tuned via contrastive ranking beat frontier LLMs like GPT-5.2 and Claude-Sonnet-4.6 by over 10 percentage points on real-world entity disambiguation.
Understanding the historical allocation and distribution of research funding advances our knowledge of how scientific research is supported across fields, institutions, and regions. However, large-scale analyses are hindered by the lack of comprehensive funder name disambiguation solutions, as funder names often exhibit spelling variations, translations, abbreviations, and inconsistent levels of granularity. In this paper, we present a framework for developing multilingual, multi-functional funder name disambiguation models and demonstrate its application to research publications in biodiversity conservation. To construct a training dataset, we integrated the Research Organization Registry (ROR), which provides unique identifiers for research organizations, with two publication datasets: the Web of Science (WoS) and the Crossref Open Funder Registry (OFR). We used multi-task learning with Contrastive Loss and Multiple Negatives Ranking Loss to fine-tune three open-weight embedding models from the Sentence Transformer, Gemma, and Qwen3 families. The best-performing models achieved accuracy above 0.90 when matching WoS funder names to ROR identifiers, outperforming general-purpose LLMs, including GPT-5.2, Claude-Sonnet-4.6, and Gemini-2.5-Flash, by more than 0.1. For funder names not indexed in ROR, we constructed a similarity network among funder names and identified clusters within it. Finally, we analyzed the disambiguation results and highlighted challenges arising from limited knowledge of smaller funders and funders from non-English-speaking countries. This work provides a reusable framework for funder name disambiguation with potential applicability across different model architectures and datasets, featuring cost-effective training data creation and multi-task learning and disambiguation.
