Database of NIH grants using machine-learned categories and graphical clustering

By E. Talley, David Newman, David Mimno, B. Herr, Hanna M. Wallach, G. Burns, A. Leenders, Andrew McCallum. Nature Methods, 2011

298 citations recorded by the source corpus.

framework that is based on scientific research rather than NIH administrative and categorical designations. We found that topic-based categories are not strictly associated with the missions of individual Institutes but instead cut across the NIH, albeit in varying proportions consistent with each Institute’s distinct mission (Supplementary Table 1). The graphical map layout (Fig. 1) shows a global research structure that is logically coherent but only loosely related to Institute organization (Supplementary Table 1). We describe four example use cases (Supplementary Data). First, we show a query using an algorithm-derived category relevant to angiogenesis (Supplementary Fig. 1). Unlike standard keywordbased searches, this type of query allows retrieval of grants that are truly focused on a particular research area. In addition, the resulting graphical clusters reveal clear patterns in the relationships between the retrieved grants and the multiple Institutes funding this research. Second, we examine an NIH peer review study section. The database categories and clusters clarify the complex relationship between the NIH Institutes and the centralized NIH peer review system, which is distinct and independent from the Institutes. Third, we show an analysis of the NIH RCDC category ‘sleep research’ in conjunction with the database topics, the latter Database of NIH grants using machine-learned categories and graphical clustering

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/fc94c48780652164f1843e86678adb4289219076

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Sociology in the Era of Big Data: The Ascent of Forensic Social Science (2015, unknown): “Machine learning is a powerful tool that has assisted companies in many domains with various engineering questions (Talley et al. 2011).”
Feuding Families and Former Friends: Unsupervised Learning for Dynamic Fictional Relationships (2016, background): “More generally, topic models and related generative models are a central tool for understanding large corpora from science (Talley et al., 2011) to politics (Nguyen et al.”
Revealing protein–lncRNA interaction (2015, methodology): “RPISeq [107], catRAPID [108], the method described in Wang et al.”
Interpretable Machine Learning : Lessons from Topic Modeling (2016, background): “For example, topic modeling research has found that domain users are unlikely to trust topic models if some of the topics look incoherent or do not meet prior expectations [16, 23].”
Applications of Topic Models (2017, background): “A different, but related problem is understanding large document collections, common in science policy [Talley et al., 2011], journalism, and the humanities [Moretti, 2013a].”
Relative Citation Ratio (RCR): A New Metric That Uses Citation Rates to Measure Influence at the Article Level (2015, background): “Because modern fields of biomedical research exist as a spectrum rather than discrete and separate fields (Talley et al., 2011), we decided to take a more nuanced approach to defining an article’s field.”
A Data-Driven Analysis of Workers' Earnings on Amazon Mechanical Turk (2017, methodology): “methods are often used to categorize a set of unlabeled documents into a finite number of coherent topic clusters [7,27,61], in turn helping researchers to navigate a large corpus (HIT documents in our case) [17,18,24,52].”
Termite: visualization techniques for assessing textual topic models (2012, background): “[24] examined 110,000 NIH grants and applied LDA to uncover 700 latent topics.”
Interpretation and trust: designing model-driven visualizations for text analysis (2012, background): “Hall et al. [25] removed 64 topics and inserted 10 hand-crafted topics in order to complete their investigation; Talley et al. [46] removed poor topics and ﬂagged questionable topics in their visualization.”