Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Scale analysis”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,207 records · Page 67
All sourcesNASA NTRSDOE OSTI
DOE OSTI2017

Decay heat uncertainty for BWR used fuel due to modeling and nuclear data uncertainties

Characterization of the energy released from radionuclide decay in nuclear fuel discharged from reactors is essential for the design, safety, and licensing analyses of used nuclear fuel storage, transportation, and repository systems. There are a limited number of decay heat measurements available for commercial used fuel applications. Because decay heat measurements can be expensive or impractical for covering the multitude of existing fuel designs, operating conditions, and specific application purposes, decay heat estimation relies heavily on computer code prediction. Uncertainty evaluation for calculated decay heat is an important aspect when assessing code prediction and a key factor supporting decision making for used fuel applications. While previous studies have largely focused on uncertainties in code predictions due to nuclear data uncertainties, this study discusses uncertainties in calculated decay heat due to uncertainties in assembly modeling parameters as well as in nuclear data. Capabilities in the SCALE nuclear analysis code system were used to quantify the effect on calculated decay heat of uncertainties in nuclear data and selected manufacturing and operation parameters for a typical boiling water reactor (BWR) fuel assembly. Furthermore, the BWR fuel assembly used as the reference case for this study was selected from a set of assemblies for which high-quality decay heat measurements are available, to assess the significance of the results through comparison with calculated and measured decay heat data.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2012

Ion induced spinodal dewetting of thin solid films

We present experimental data and numerical simulations in order to show that the mechanism of spinodal dewetting is active during ion beam irradiation of thin solid films. The expected scaling law for the characteristic wavelengths versus the initial film thickness is modified by the presence of sputtering. The conclusion is fully supported by model simulation which shows a square law dependence for null sputtering yield and a bimodal trend when sputtering is included. This result is in contrast to earlier studies and opens the possibility to control and use ion induced dewetting for the fabrication of functional nanostructures.

36 MATERIALS SCIENCE↗
DOE OSTI2023

DESI mock challenge: Halo and galaxy catalogues with the bias assignment method

We present a novel approach to the construction of mock galaxy catalogues for large-scale structure analysis based on the distribution of dark matter halos obtained with effective bias models at the field level. We aim to produce mock galaxy catalogues capable of generating accurate covariance matrices for a number of cosmological probes that are expected to be measured in current and forthcoming galaxy redshift surveys (e.g. two- and three-point statistics). The construction of the catalogues shown in this paper is part of a mock-comparison project within the Dark Energy Spectroscopic Instrument (DESI) collaboration. We use the bias assignment method ( BAM ) to model the statistics of halo distribution through a learning algorithm using a few detailed N-body simulations, and approximated gravity solvers based on Lagrangian perturbation theory. We introduce cosmic-web-dependent corrections to modelling redshift-space distortions at the N-body level – both in the halo and galaxy distributions –, as well as a multi-scale approach for accurate assignment of halo properties. Using specific models of halo occupation distributions to populate halos, we generate galaxy mocks with the expected number density and central-satellite fraction of emission-line galaxies, which are a key target of the DESI experiment. BAM generates mock catalogues with per cent accuracy in a number of summary statistics, such as the abundance, the two- and three-point statistics of halo distributions, both in real and redshift space. In particular, the mock galaxy catalogues display ~3%-10% accuracy in the multipoles of the power spectrum up to scales of k ~ 0.4 h -1 Mpc. We show that covariance matrices of two- and three-point statistics obtained with BAM display a similar structure to the reference simulation. BAM offers an efficient way to produce mock halo catalogues with accurate two- and three-point statistics and is able to generate a variety of multi-tracer catalogues with precise covariance matrices of several cosmological probes. We discuss future developments of the algorithm towards mock production in DESI and other galaxy-redshift surveys.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

Electrostatic confinement and manipulation of DNA molecules for genome analysis

Very large DNA molecules enable comprehensive analysis of complex genomes, such as human, cancer, and plants because they span across sequence repeats and complex somatic events. When physically manipulated, or analyzed as single molecules, long polyelectrolytes are problematic because of mechanical considerations that include shear-mediated breakage, dealing with the massive size of these coils, or the length of stretched DNAs using common experimental techniques and fluidic devices. Accordingly, we harness analyte “issues” as exploitable advantages by our invention and characterization of the “molecular gate,” which controls and synchronizes formation of stretched DNA molecules as DNA dumbbells within nanoslit geometries. Molecular gate geometries comprise micro- and nanoscale features designed to synergize very low ionic strength conditions in ways we show effectively create an “electrostatic bottle.” This effect greatly enhances molecular confinement within large slit geometries and supports facile, synchronized electrokinetic loading of nanoslits, even without dumbbell formation. Device geometries were considered at the molecular and continuum scales through computer simulations, which also guided our efforts to optimize design and functionalities. In addition, we show that the molecular gate may govern DNA separations because DNA molecules can be electrokinetically triggered, by varying applied voltage, to enter slits in a size-dependent manner. Lastly, mapping the Mesoplasma florum genome, via synchronized dumbbell formation, validates our nascent approach as a viable starting point for advanced development that will build an integrated system capable of large-scale genome analysis.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2016

Incremental k-core decomposition: Algorithms and evaluation

A k-core of a graph is a maximal connected subgraph in which every vertex is connected to at least k vertices in the subgraph. k-core decomposition is often used in large-scale network analysis, such as community detection, protein function prediction, visualization, and solving NP-hard problems on real networks efficiently, like maximal clique finding. In many real-world applications, networks change over time. As a result, it is essential to develop efficient incremental algorithms for dynamic graph data. In this paper, we propose a suite of incremental k-core decomposition algorithms for dynamic graph data. These algorithms locate a small subgraph that is guaranteed to contain the list of vertices whose maximum k-core values have changed and efficiently process this subgraph to update the k-core decomposition. We present incremental algorithms for both insertion and deletion operations, and propose auxiliary vertex state maintenance techniques that can further accelerate these operations. Our results show a significant reduction in runtime compared to non-incremental alternatives. We illustrate the efficiency of our algorithms on different types of real and synthetic graphs, at varying scales. Furthermore, for a graph of 16 million vertices, we observe relative throughputs reaching a million times, relative to the non-incremental algorithms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Scaling laws and stability of nano-sized defect clusters in niobium via atomistic simulations and statistical analysis

The predictions of scaling laws for the structure and properties of defect clusters are generally limited to small defect clusters in their ground-state configurations. We investigated the size and geometrical configuration dependence of nano-sized defect clusters in niobium (Nb) using molecular dynamics. We studied the structure and stability of large clusters of size up to fifty defects for vacancies and one hundred defects for interstitials, as well as the role of helium and metastable configurations on the stability of these clusters. We compared three different interatomic potentials in order to determine the relative stability of these clusters as a function of their size and geometrical configurations. Additionally, we conducted a statistical analysis to predict the formation and binding energies of interstitial clusters as a function of both their size and configuration. We find that the size dependence of vacancy and interstitial clusters can be approximated by functional forms that account for bulk and surface effects as well as some considerations of elastic interactions. We also find that helium and metastable configurations can make vacancy and interstitial clusters thermally stable depending on the configuration. Our parameterized functional forms for the formation and binding energies are valid for a very broad range of defect sizes and configurations making it possible to be used directly in a coarse-grained modeling strategy such as Monte Carlo, cluster dynamics or dislocation dynamics which look at defect accumulation and evolution in microstructures.

36 MATERIALS SCIENCE↗
DOE OSTI2013

Extending the subspace hybrid method for eigenvalue problems in reactor physics calculation

This paper presents an innovative hybrid Monte-Carlo-Deterministic method denoted by the SUBSPACE method designed for improving the efficiency of hybrid methods for reactor analysis applications. The SUBSPACE method achieves its high computational efficiency by taking advantage of the existing correlations between desired responses. Recently, significant gains in computational efficiency have been demonstrated using this method for source driven problems. Within this work the mathematical theory behind the SUBSPACE method is introduced and extended to address core wide level k-eigenvalue problems. The method's efficiency is demonstrated based on a three-dimensional quarter-core problem, where responses are sought on the pin cell level. The SUBSPACE method is compared to the FW-CADIS method and is found to be more efficient for the utilized test problem because of the reason that the FW-CADIS method solves a forward eigenvalue problem and an adjoint fixed-source problem while the SUBSPACE method only solves an adjoint fixed-source problem. Based on the favorable results obtained here, we are confident that the applicability of Monte Carlo for large scale reactor analysis could be realized in the near future. (authors)

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Comparing cross-platform performance via node-to-node scaling studies

Due to the increasing diversity of high-performance computing architectures, researchers and practitioners are increasingly interested in comparing a code’s performance and scalability across different platforms. However, there is a lack of available guidance on how to actually set up and analyze such cross-platform studies. In this paper, we contend that the natural base unit of computing for such studies is a single compute node on each platform and offer guidance in setting up, running, and analyzing node-to-node scaling studies. In conclusion, we propose templates for presenting scaling results of these studies and provide several case studies highlighting the benefits of this approach.

cross-platform studies↗
DOE OSTI2012

The U.S. Forest Service's analysis of cumulative effects to wildlife: A study of legal standards, current practice, and ongoing challenges on a National Forest

Cumulative effects analysis (CEA) allows natural resource managers to understand the status of resources in historical context, learn from past management actions, and adapt future activities accordingly. U.S. federal agencies are required to complete CEA as part of environmental impact assessment under the National Environmental Policy Act (NEPA). Past research on CEA as part of NEPA has identified significant deficiencies in CEA practice, suggested methodologies for handling difficult aspects of CEA, and analyzed the rise in litigation over CEA in U.S. courts. This article provides a review of the literature and legal standards related to CEA as it is done under NEPA and then examines current practice on a U.S. National Forest, utilizing qualitative methods in order to provide a detailed understanding of current approaches to CEA. Research objectives were to understand current practice, investigate ongoing challenges, and identify impediments to improvement. Methods included a systematic review of a set of NEPA documents and semi-structured interviews with practitioners, scientists, and members of the public. Findings indicate that the primary challenges associated with CEA include: issues of both geographic and temporal scale of analysis, confusion over the purpose of the requirement, the lack of monitoring data, and problems coordinating and disseminating data. Improved monitoring strategies and programmatic analyses could support improved CEA practice.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2017

What does fault tolerant Deep Learning need from MPI?

Deep Learning (DL) algorithms have become the {\em de facto} Machine Learning (ML) algorithm for large scale data analysis. DL algorithms are computationally expensive -- even distributed DL implementations which use MPI require days of training (model learning) time on commonly studied datasets. Long running DL applications become susceptible to faults -- requiring development of a fault tolerant system infrastructure, in addition to fault tolerant DL algorithms. This raises an important question: {\em What is needed from MPI for designing fault tolerant DL implementations?} In this paper, we address this problem for permanent faults. We motivate the need for a fault tolerant MPI specification by an in-depth consideration of recent innovations in DL algorithms and their properties, which drive the need for specific fault tolerance features. We present an in-depth discussion on the suitability of different parallelism types (model, data and hybrid); a need (or lack thereof) for check-pointing of any critical data structures; and most importantly, consideration for several fault tolerance proposals (user-level fault mitigation (ULFM), Reinit) in MPI and their applicability to fault tolerant DL implementations. We leverage a distributed memory implementation of Caffe, currently available under the Machine Learning Toolkit for Extreme Scale (MaTEx). We implement our approaches by extending MaTEx-Caffe for using ULFM-based implementation. Our evaluation using the ImageNet dataset and AlexNet neural network topology demonstrates the effectiveness of the proposed fault tolerant DL implementation using OpenMPI based ULFM.

Amatya, Vinay C.↗
DOE OSTI2020

Assessing CLUBB PDF Closure Assumptions for a Continental Shallow-to-Deep Convective Transition Case Over Multiple Spatial Scales

Assumed-PDF (probability density function) higher-order turbulence closures (APHOCs) are now widely used for parameterizing boundary layer turbulence and shallow convection in Earth system models (ESMs). A better understanding of the resolution-dependent behavior of APHOCs is essential for improving the performance of next-generation ESMs with intended horizontal resolutions finer than 10 km. In this study, we evaluate the PDF family of Analytic double-Gaussian 1 implemented in Cloud Layers Unified By Binormals (CLUBB) over a range of spatial scales (Dx) from 2 to 100 km. A 120-km-wide large eddy simulation (LES) for a continental convection case during 2016 Holistic Interactions of Shallow Clouds, Aerosols, and Land-Ecosystems (HI-SCALE) field campaign serves as benchmark to evaluate the PDF closure using an off-line approach. We find during the shallow convection period, the CLUBB PDF closure tends to produce positive biases of cloud properties and liquid water flux near cloud base for all scales of analysis. It produces negative biases for these variables near cloud top that are more severe for Dx larger than 25 km. Results show that replacing the CLUBB-parameterized moisture and temperature skewnesses with LES-derived ones can fix most of the biases if clipping of input moments is allowed to prevent the occurrence of unrealizable solutions. Overall, the performance of the PDF closure is better for smaller Dx = 2–5 km than for larger Dx = 50–100 km; for a given grid spacing, it is better when the convective clouds become deeper in the late afternoon. Likely causes for the resolution dependence and implications for improving the PDF closure are discussed.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

AutoBEM: A scalable framework for nationwide building energy simulation and retrofit evaluation in the United States

This paper presents AutoBEM, an integrated, automated framework for nationwide building energy modeling and retrofit evaluation in the United States. Unlike prior UBEM platforms that either rely primarily on representative stock sampling or operate at city scale, AutoBEM automates the generation of building-resolved, physics-based EnergyPlus/OpenStudio simulation models at national scale using GIS-derived geometry, prototype-based assumptions, and standardized scalable workflows. Leveraging the Model America dataset and high-performance computing, AutoBEM generates and simulates energy models for 122.9 million buildings, representing 97.8% of the U.S. building stock. These models are being made publicly and freely available as the Model America v1.0 (MAv1) dataset. AutoBEM supports detailed, building-level assessments of energy consumption, CO2 emissions, and post-processed anthropogenic heat emissions (AHE), and evaluates 151 energy conservation measures (ECMs) using localized utility pricing and building characteristics. In addition, AutoBEM incorporates both typical and future climate conditions through integration with Typical Meteorological Year (TMY) and Future TMY (fTMY) weather data derived from IPCC scenarios. In a case study of Phoenix, Arizona, AutoBEM identified several high-efficiency HVAC upgrades and selected envelope measures with short modeled payback periods (1.5 years) for certain building types and standards. Simulations under future climate scenarios (SSP5–RCP8.5) project an 11.3% increase in electricity use and a 32% reduction in natural gas demand by 2100, underscoring the need for climate-adaptive retrofit planning. By enabling reproducible, bottom-up, and location-specific analysis at scale, AutoBEM provides a step toward a national digital twin of the built environment and supports data-driven screening and planning for decarbonization, resilience, and energy equity.

Li, Hang [ORNL] (ORCID:0000000306001920)↗
DOE OSTI2013

Mathematical and Computational Tools for Predictive Simulation of Complex Coupled Systems under Uncertainty

Methods and algorithms are developed to enable the accurate analysis of problems that exhibit interacting physical processes with uncertainties. These uncertainties can pertain either to each of the physical processes or to the manner in which they depend on each others. These problems are cast within a polynomial chaos framework and their solution then involves either solving a large system of algebraic equations or a high dimensional numerical quadrature. In both cases, the curse of dimensionality is manifested. Procedures are developed for the efficient evaluation of the resulting linear equations that advantage of the block sparse structure of these equations, resulting in a block recursive Schur complement construction. In addition, embedded quadratures are constructed that permit the evaluation of very high-dimensional integrals using low-dimensional quadratures adapted to particular quantities of interest. The low-dimensional integration is carried out in a transformed measure space in which the quantity of interest is low-dimensional. Finally, a procedure is also developed to discover a low-dimensional manifold, embedded in the initial high-dimensional one, in which scalar quantities of interest exist. This approach permits the functional expression of the reduced space in terms of the original space, thus permitting cross-scale sensitivity analysis.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

The effect of sanitary landfill leachate aging on the biological treatment and assessment of photoelectrooxidation as a pre-treatment process

Highlights: • Photoelectrooxidation (PEO) emerges as a new technology for leachate treatment. • Aging of sanitary landfills acts on leachate composition and biodegradability. • PEO is applied as leachate pretreatment before the biological processes. • PEO produced significant changes in the leachate matrix, easing biological process. - Abstract: The sanitary landfill leachate is a dark liquid, of highly variable composition, with recalcitrant features that hamper conventional biological treatment. The physical–chemical characteristics of the leachate along the landfill aging, as well as their effects on the efficiency of the conventional treatment, were evaluated at this paper. The feasibility of photoelectrooxidation process as an alternative technique for treatment of landfill leachates was also determined. Photoelectrooxidation experiments were conducted in a bench-scale reactor. Analysis of the raw leachate revealed many critical parameters demonstrating that the recalcitrance of leachate tends to increase with time, directly influencing the decline in efficiency of the conventional treatment currently employed. The effects of current density and lamp power were investigated. Using a 400 W power lamp and a current density of 31.5 mA cm{sup −2}, 53% and 61% efficiency for the removal of ammoniacal nitrogen and chemical oxygen demand were respectively achieved by applying photoelectrooxidation process. With the removal of these pollutants, downstream biological treatment should be improved. These results demonstrate that photoelectrooxidation is a feasible technique for the treatment of sanitary landfill leachate, even considering this effluent’s high resistance to treatment.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS2020

Multibody Based Digital Astronaut Dynamics Simulation

BACKGROUND: This study provides the Software, Robotics, & Simulation Division at the NASA Johnson Space Center with a verification tool for multibody dynamics simulation requiring human motion. The motivation stems from current studies of several Vibration Isolation & Stabilization(VIS)system designs that attenuate the moments and forces which would be transmitted to a spacecraft during an exercise. A multibody dynamics model for a proposed VIS was available previously[1], therefore modeling of the VIS was not needed for this work. The interest here is in creating the multibody dynamics model of an astronaut in motion which may be utilized independently or while attached to a mechanism. An existing simulation [2] that utilizes OpenSim [3,4] and an in-house multibody dynamics package (MBDyn) [5] is used in order to verify the astronaut model. The main advantage this model will have over the existing simulation is that everything will be processed in one tool. METHODS AND RESULTS: Creating the simulation required; estimation of Body Segment Inertial Parameters (BSIP),a multibody model of the human-VIS system, joint acceleration profiles, and input files for MBDyn, which is used for this analysis. The scaling factors provided by Dumas et al. [6] are utilized in estimating the BSIP. Anthropometric data are used for estimating these parameters, the Anthropometric Survey of US Army Personnel (ANSUR II) [7] was the source. The astronaut model consists of 15 bodies, 14 joints and 32 degrees of freedom, with the dynamics topology generated using the center of mass locations and anthropometric data. MBDyn has an option for prescribed joint motion (PJM), which requires joint acceleration data as input. The joint angle data is first obtained from a motion capture system and then processed through code that has been created to generate approximate joint acceleration profiles. The topology tree for the astronaut model begins at the right foot up to the pelvis where there is one branch for going down the left leg and another for the torso. The torso branch leads to branches for the arms and a leaf body for the head/neck segment. For attachment to the VIS, the heel of the right foot is connected to the VIS platform through a fixed joint, resembling a foot restraint. The left foot does not attach to the platform in order to prevent a system with a closed loop. Topology and symmetry of the astronaut model were verified through kinematic analysis. Further verification of the forces and moments transmitted to the VIS were verified against the existing simulation. There was a satisfactory level of agreement when testing a simple motion, for example, rocking back and forth. Full exercise motions are to be tested soon. The main outcome has been a novel application of MBDyn for biomechanics modeling that is now available for dynamic simulations involving human motion. The estimation of BSIP was another useful result of this study, requiring only 15 inputs for generating mass properties of a theoretical astronaut model. Expansion on this work is possible by going through an alternative route in obtaining the joint motion data. Instead of high-tech and often expensive motion capture systems, an individual may watch videos with high focus and at a slow motion for each individual segment in order to determine the initial and final time and angle for that specific degree of freedom. Synthetic trajectories may also be created if there is no video reference available.

F N Matari↗
DOE OSTI2020

Automated pipeline framework for processing of large-scale building energy time series data

Commercial buildings account for one third of the total electricity consumption in the United States and a significant amount of this energy is wasted. Therefore, there is a need for “virtual” energy audits, to identify energy inefficiencies and their associated savings opportunities using methods that can be non-intrusive and automated for application to large populations of buildings. Here we demonstrate virtual energy audits applied to large populations of buildings’ time-series smart-meter data using a systematic approach and a fully automated Building Energy Analytics (BEA) Pipeline that unifies, cleans, stores and analyzes building energy datasets in a non-relational data warehouse for efficient insights and results. This BEA pipeline is based on a custom compute job scheduler for a high performance computing cluster to enable parallel processing of Slurm jobs. Within the analytics pipeline, we introduced a data qualification tool that enhances data quality by fixing common errors, while also detecting abnormalities in a building’s daily operation using hierarchical clustering. We analyze the HVAC scheduling of a population of 816 buildings, using this analytics pipeline, as part of a cross-sectional study. With our approach, this sample of 816 buildings is improved in data quality and is efficiently analyzed in 34 minutes, which is 85 times faster than the time taken by a sequential processing. The analytical results for the HVAC operational hours of these buildings show that among 10 building use types, food sales buildings with 17.75 hours of daily HVAC cooling operation are decent targets for HVAC savings. Overall, this analytics pipeline enables the identification of statistically significant results from population based studies of large numbers of building energy time-series datasets with robust results. These types of BEA studies can explore numerous factors impacting building energy efficiency and virtual building energy audits. This approach enables a new generation of data-driven buildings energy analysis at scale.

36 MATERIALS SCIENCE↗
DOE OSTI2017

On mechanical waves and Doppler shifts from moving boundaries

We investigate the propagation of infinitesimal harmonic mechanical waves emitted from a boundary with variable velocity and arriving at a stationary observer. In the classical Doppler effect, X s (t)=vt is the location of the source with constant velocity v. In the present work, however, we consider a source co-located with a moving boundary x=X s (t), where X s (t) can have an arbitrary functional form. For ‘slowly moving’ boundaries (i.e., ones for which the timescale set by the mechanical motion is large in comparison to the inverse of the frequency of the emitted wave), we present a multiple-scale asymptotic analysis of the moving boundary problem for the linear wave equation. Here, we obtain a closed-form leading-order (with respect to the latter small parameter) solution and show that the variable velocity of the boundary results not only in frequency modulation but also in amplitude modulation of the received signal. Consequently, our results extend the applicability of two basic tenets of the theory of a moving source on a stationary domain, specifically that (i) $.\atop{x}_s$ for non-uniform boundary motion can be inserted in place of the constant velocity v in the classical Doppler formula and (ii) that the non-uniform boundary motion introduces variability in the amplitude of the wave. The specific examples of decelerating and oscillatory boundary motion are worked out and illustrated.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

AI-powered topic modeling: comparing LDA and BERTopic in analyzing opioid-related cardiovascular risks in women

Topic modeling is a crucial technique in natural language processing (NLP), enabling the extraction of latent themes from large text corpora. Traditional topic modeling, such as Latent Dirichlet Allocation (LDA), faces limitations in capturing the semantic relationships in the text document although it has been widely applied in text mining. BERTopic, created in 2022, leveraged advances in deep learning and can capture the contextual relationships between words. In this work, we integrated Artificial Intelligence (AI) modules to LDA and BERTopic and provided a comprehensive comparison on the analysis of prescription opioid-related cardiovascular risks in women. Opioid use can increase the risk of cardiovascular problems in women such as arrhythmia, hypotension etc. 1,837 abstracts were retrieved and downloaded from PubMed as of April 2024 using three Medical Subject Headings (MeSH) words: “opioid,” “cardiovascular,” and “women.” Machine Learning of Language Toolkit (MALLET) was employed for the implementation of LDA. BioBERT was used for document embedding in BERTopic. Eighteen was selected as the optimal topic number for MALLET and 23 for BERTopic. ChatGPT-4-Turbo was integrated to interpret and compare the results. The short descriptions created by ChatGPT for each topic from LDA and BERTopic were highly correlated, and the performance accuracies of LDA and BERTopic were similar as determined by expert manual reviews of the abstracts grouped by their predominant topics. The results of the t-SNE (t-distributed Stochastic Neighbor Embedding) plots showed that the clusters created from BERTopic were more compact and well-separated, representing improved coherence and distinctiveness between the topics. Our findings indicated that AI algorithms could augment both traditional and contemporary topic modeling techniques. In addition, BERTopic has the connection port for ChatGPT-4-Turbo or other large language models in its algorithm for automatic interpretation, while with LDA interpretation must be manually, and needs special procedures for data pre-processing and stop words exclusion. Therefore, while LDA remains valuable for large-scale text analysis with resource constraints, AI-assisted BERTopic offers significant advantages in providing the enhanced interpretability and the improved semantic coherence for extracting valuable insights from textual data.

Research & Experimental Medicine↗