AuraTracer智迹闻
中文

EVENT DOSSIER

Measuring the Novelty of Biomedical Papers Using the Latent Distances between Knowledge Units

2026-09-07 12:00 Science across 2 days 🔥 47.2 heat score
2sources
2days unfolding
47.2heat score
2mentions
SummaryAI generated

The researchers proposed a comprehensive measurement method for novelty, aiming to quantify potential distances by integrating the networks, semantics, and hierarchical relationships between knowledge units. The study was analyzed and validated based on 142,036 biomedical papers published in PLoS ONE and the validation dataset from the H1 Connect platform. The results showed that the three types of relationships respectively captured different characteristics of potential distances between MeSH terms; compared to the metrics proposed by Uzzi et al. (2013), this measurement method demonstrated stronger consistency with peer review results; moreover, combining all three distance metrics was more effective at identifying novel papers than a single perspective.

Related eventsRELATED EVENTS
Key entitiesKEY ENTITIES
H1 ConnectPLoS ONE

Coverage · reports per dayLANGUAGE SPLIT

Entity relations
H1 Connect × PLoS ONE2

Integrated timelineUNIFIED TIMELINE

  1. 2026-09-04

    Measuring the Novelty of Biomedical Pap…

    本研究提出一种综合新颖度测量方法,通过整合知识单元间的网络、语义及层级三种关系来量化潜在距离。研究基于 PLoS ONE 发表的 142,036 篇文章及 H1 Connect 平台验证数据集进行验证,结果显示:每种关系类型捕获了 MeS…

  2. 2026-09-07

    Measuring the Novelty of Biomedical Pap…

    本研究提出一种综合新颖性测量方法,通过整合知识单元间的网络、语义及层级关系来量化潜在距离。研究基于 PLoS ONE 发表的 142,036 篇文章及 H1 Connect 平台的验证数据集进行分析。结果显示,三种关系类型分别捕捉了 Me…

SignalsSIGNALS

Keyword heat
  • PLoS ONE2
  • H1 Connect2

All reports (2)SOURCES

A arXiv cs.CL en 2026-09-04 22:14

Measuring the Novelty of Biomedical Papers Using the Latent Distances between Knowledge Units

本研究提出一种综合新颖度测量方法,通过整合知识单元间的网络、语义及层级三种关系来量化潜在距离。研究基于 PLoS ONE 发表的 142,036 篇文章及 H1 Connect 平台验证数据集进行验证,结果显示:每种关系类型捕获了 MeSH 术语间不同的潜在距离;与 Uzzi 等人(2013)提出的常用指标相比,该测量与同行判断的一致性更强;且结合三种距离指标比单一视角更能有效识别新颖论文。

A arXiv cs.CL en 2026-09-07 12:00

Measuring the Novelty of Biomedical Papers Using the Latent Distances between Knowledge Units

本研究提出一种综合新颖性测量方法,通过整合知识单元间的网络、语义及层级关系来量化潜在距离。研究基于 PLoS ONE 发表的 142,036 篇文章及 H1 Connect 平台的验证数据集进行分析。结果显示,三种关系类型分别捕捉了 MeSH 术语间不同的潜在距离;相较于 Uzzi 等人(2013)提出的指标,该测量与同行评议的一致性更强;且结合所有三种距离指标能比单一视角更有效地识别新颖论文。