ConceptDrift: Leveraging Spatial, Temporal and Semantic Evolution of Biomedical Concepts for Hypothesis Generation

Amir Hassan Shariatmadari; Alireza Jafari; Sikun Guo; Sneha Srinivasan; Nathan C Sheffield; Aidong Zhang; Kishlay Jha

doi:10.1093/bioinformatics/btaf563

Back

ConceptDrift: Leveraging Spatial, Temporal and Semantic Evolution of Biomedical Concepts for Hypothesis Generation

Journal article

Open access

Peer reviewed

ConceptDrift: Leveraging Spatial, Temporal and Semantic Evolution of Biomedical Concepts for Hypothesis Generation

Amir Hassan Shariatmadari, Alireza Jafari, Sikun Guo, Sneha Srinivasan, Nathan C Sheffield, Aidong Zhang and Kishlay Jha

Bioinformatics (Oxford, England), Vol.41(11), btaf563

11/2025

DOI: 10.1093/bioinformatics/btaf563

PMCID: PMC12582365

PMID: 41148039

Files and links (1)

url

https://doi.org/10.1093/bioinformatics/btaf563View

Published (Version of record) Open Access

Abstract

Hypothesis generation is a fundamental problem in biomedical text mining that aims to generate ideas that are new, interesting, and plausible by discovering unexplored links between biomedical concepts. Despite significant advances made by existing approaches, they do not fully leverage the evolutionary properties of biomedical concepts. This is limiting because scientific knowledge continually evolves over time, with new facts being added and old ones becoming obsolete. Thus, it is crucial to capture the evolutionary properties of biomedical concepts from multiple perspectives (e.g., spatial, temporal, and semantic) to generate hypotheses that reflect the up-to-date information landscape of the biomedical domain. We introduce a novel framework, ConceptDrift, that models the hypothesis generation task as a sequence of temporal graphlets and simultaneously encodes spatial, temporal, and semantic change. Unlike existing approaches that treat these dimensions independently, ConceptDrift is the first to provide a holistic understanding of concept evolution by integrating them into a unified framework. Grounded in the theories of the Distributional Hypothesis and Conceptual Change, our method adapts these principles to the unique challenges of large-scale biomedical literature. We conduct extensive experiments across multiple datasets and demonstrate that ConceptDrift consistently outperforms state-of-the-art baselines in generating accurate and meaningful hypotheses. Our framework shows immediate practical benefits for web-based literature mining tools in life sciences and biomedicine, offering more robust and predictive feature representations. https://github.com/amir-hassan25/ConceptDrift (DOI: 10.6084/m9.figshare.29975476).

Data Mining

Temporal Graph Learning

Hypothesis Generation

Details

Title: Subtitle: ConceptDrift: Leveraging Spatial, Temporal and Semantic Evolution of Biomedical Concepts for Hypothesis Generation
Creators: Amir Hassan Shariatmadari - University of Virginia
Alireza Jafari - University of Virginia
Sikun Guo - University of Virginia
Sneha Srinivasan - University of Virginia
Nathan C Sheffield - University of Virginia
Aidong Zhang - University of Virginia
Kishlay Jha - University of Iowa
Resource Type: Journal article
Publication Details: Bioinformatics (Oxford, England), Vol.41(11), btaf563
DOI: 10.1093/bioinformatics/btaf563
PMID: 41148039
PMCID: PMC12582365
NLM abbreviation: Bioinformatics
ISSN: 1367-4811
eISSN: 1367-4811
Publisher: Oxford University Press
Grant note: US National Institute of Health (NIH)National Science Foundation (NSF): R01LM014012-01A1, IIS-2106913, BIO-2313865, SCH-2500341, SCH-2500344
This work is supported in part by the US National Institute of Health (NIH) and the National Science Foundation (NSF) under grants R01LM014012-01A1, IIS-2106913, BIO-2313865, SCH-2500341 and SCH-2500344.
Language: English
Electronic publication date: 10/28/2025
Date published: 11/2025
Academic Unit: Electrical and Computer Engineering
Record Identifier: 9985019032302771

Metrics

9 Record Views