Users of topic modeling methods often have knowledge about the composition of words that should have high or low probability in various topics. We incorporate such domain knowledge using a novel Dirichlet Forest prior in a Latent Dirichlet Allocation framework. The prior is a mixture of Dirichlet tree distributions with special structures. We present its construction, and inference via collapsed Gibbs sampling. Experiments on synthetic and real datasets demonstrate our model's ability to follow and generalize beyond user-specified domain knowledge.
Download full-text PDF |
Source |
|---|---|
| http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2943854 | PMC |
| http://dx.doi.org/10.1145/1553374.1553378 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Examining the associations between social support and postpartum depression symptoms among adolescent mothers in Nairobi, Kenya.
Reprod Health
January 2025
Sexual, Reproductive, Maternal, Newborn, Child and Adolescent Health (SRMNCAH) Unit, African Population and Health Research Center, Nairobi, Kenya.
Background: Globally, adolescent mothers are at increased risk for postpartum depression (PPD). In Kenya, 15% of adolescent girls become mothers before the age of 18. While social support can buffer a mother's risk of PPD, there are gaps in knowledge as to whether-and which types-of social support are protective for adolescent mothers in Kenya.
View Article and Find Full Text PDFIntroduction: Healthcare provider burnout is highly prevalent and has negative consequences. However, many healthcare workers in LMICs, including Nepal, rarely recognize or ameliorate it. This problem is worse in rural settings.
View Article and Find Full Text PDFRole of inflammasomes in acute respiratory distress syndrome.
Thorax
January 2025
Victor Phillip Dahdaleh Heart & Lung Research Institute, University of Cambridge, Cambridge, UK.
Acute respiratory distress syndrome (ARDS) is present in >10% of all people admitted to critical care and is associated with severe morbidity and mortality. Despite more than half a century since its first description, no efficacious pharmacological therapies have been developed, and little progress has been made in improving clinical outcomes. Neutrophils are the principal drivers of ARDS, with their priming and subsequent aberrant downstream functions, including interleukin (IL) 1β and IL-18 secretion, central to the disease pathogenesis.
View Article and Find Full Text PDFEHM: Exploring dynamic alignment and hierarchical clustering in unsupervised domain adaptation via high-order moment-guided contrastive learning.
Neural Netw
January 2025
College of Information Science & Electronic Engineering, Zhejiang University, Hangzhou, 310027, China. Electronic address:
Unsupervised domain adaptation (UDA) aims to annotate unlabeled target domain samples using transferable knowledge learned from the labeled source domain. Optimal transport (OT) is a widely adopted probability metric in transfer learning for quantifying domain discrepancy. However, many existing OT-based UDA methods usually employ an entropic regularization term to solve the OT optimization problem, inevitably resulting in a biased estimation of domain discrepancy.
View Article and Find Full Text PDFEvolution of Artificial Intelligence in Medical Education From 2000 to 2024: Bibliometric Analysis.
Interact J Med Res
January 2025
Department of Obstetrics and Gynecology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.
Background: Incorporating artificial intelligence (AI) into medical education has gained significant attention for its potential to enhance teaching and learning outcomes. However, it lacks a comprehensive study depicting the academic performance and status of AI in the medical education domain.
Objective: This study aims to analyze the social patterns, productive contributors, knowledge structure, and clusters since the 21st century.
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!