Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “BERT”

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 55 records · Page 3

Exploring Semantic Search Capability of Graph Convolutions Over a Knowledge Graph Built Using Earth Science Corpora

Traditional knowledge graphs tend to be too generic, and often perform poorly on complex scientific queries. Often times, precedence is given to pop culture over scientific knowledge for queries. This is predominantly due to the use of internet sources for building the knowledge graph. With this work, we aim to explore the effectiveness of combining a knowledge graph generated from earth science corpora with a language model and graph convolutions for the purpose of surfacing latent and related sentences given a natural language query. In this model, sentences are conceptualized in the graph as nodes which are connected through entities—words and phrases of interest found in the text—extracted using Google Cloud’s entity extraction model. The language model we used for this is Bidirectional Encoder Representations from Transformers (BERT).The sentences are given a numeric representation by the BERT model. Graph convolutions are then applied to sentence embeddings in order to obtain a vector representation of the sentence as well as the surrounding graph structure, thereby leveraging the power of adjacency inherently encoded in graph structures. With this presentation, we demonstrate the ability of graph convolutions and their improved ability to surface relevant, latent information based on the subject of the input query.

Muthukumaran Ramasubramanian↗

Classification of Notices to Airmen using Natural Language Processing

This paper establishes the feasibility of using Natural Language Processing (NLP) to classify NOTAMs or Notices to Airmen – a pilot messaging framework to gather real-time situational awareness. Present day air mobility operations heavily rely on NOTAMs. However, pilots often have difficulty interpreting NOTAMs due to the sheer volume of inapplicable messages and unclear abbreviations. Using NLP, the presented study analyzes the accuracy of classifying NOTAMs and, thereby, the efficiency of generating actionable interpretations in real time. To this effect, efficacies of four NLP neural network architectures were analyzed, including three Recurrent Neural Networks (RNNs) with GloVe, Word2Vec, and FastText word embeddings, and one trained Bi-Directional Encoder Representations from Transformers (BERT) model. The four neural networks were trained and evaluated on three open-source datasets of varying text lengths, vocabularies, and grammars, taken from e-commerce product descriptions, social media tweets, and unstructured descriptions for data and analytics services on open data marketplaces such as NASA’s Data and Reasoning Fabric (DRF) platform. This provided cross-analysis of each neural network architecture’s performance per text type. The best performing architecture, BERT, was then fine-tuned on a collection of open-source NOTAM data. Post-training, a real-time NOTAM classification service was implemented to draw inference on new NOTAMs using the trained model, which demonstrated close to 99% accuracy in classification. This modular classification service is envisioned to be integrated with a data and analytics delivery platform, such as the DRF, thus availing real-time contextualization of NOTAMs to air mobility clients, humans, and machines for enhanced decision making.

Aiden C. Szeto↗

Classification of Notices to Airmen using Natural Language Processing

This paper establishes the feasibility of using Natural Language Processing (NLP) to classify NOTAMs or Notices to Airmen – a pilot messaging framework to gather real-time situational awareness. Present day air mobility operations heavily rely on NOTAMs. However, pilots often have difficulty interpreting NOTAMs due to the sheer volume of inapplicable messages and unclear abbreviations. Using NLP, the presented study analyzes the accuracy of classifying NOTAMs and, thereby, the efficiency of generating actionable interpretations in real time. To this effect, efficacies of four NLP neural network architectures were analyzed, including three Recurrent Neural Networks (RNNs) with GloVe, Word2Vec, and FastText word embeddings, and one trained Bi-Directional Encoder Representations from Transformers (BERT) model. The four neural networks were trained and evaluated on three open-source datasets of varying text lengths, vocabularies, and grammars, taken from e-commerce product descriptions, social media tweets, and unstructured descriptions for data and analytics services on open data marketplaces such as NASA’s Data and Reasoning Fabric (DRF) platform. This provided cross-analysis of each neural network architecture’s performance per text type. The best performing architecture, BERT, was then fine-tuned on a collection of open-source NOTAM data. Post-training, a real-time NOTAM classification service was implemented to draw inference on new NOTAMs using the trained model, which demonstrated close to 99% accuracy in classification. This modular classification service is envisioned to be integrated with a data and analytics delivery platform, such as the DRF, thus availing real-time contextualization of NOTAMs to air mobility clients, humans, and machines for enhanced decision making.

Aiden Szeto↗

A survey of techniques for optimizing transformer inference

Recent years have seen a phenomenal rise in the performance and applications of transformer neural networks. The family of transformer networks, including Bidirectional Encoder Representations from Transformer (BERT), Generative Pretrained Transformer (GPT) and Vision Transformer (ViT), have shown their effectiveness across Natural Language Processing (NLP) and Computer Vision (CV) domains. Transformer-based networks such as ChatGPT have impacted the lives of common men. However, the quest for high predictive performance has led to an exponential increase in transformers' memory and compute footprint. Researchers have proposed techniques to optimize transformer inference at all levels of abstraction. Further, this paper presents a comprehensive survey of techniques for optimizing the inference phase of transformer networks. We survey techniques such as knowledge distillation, pruning, quantization, neural architecture search and lightweight network design at the algorithmic level. We further review hardware-level optimization techniques and the design of novel hardware accelerators for transformers. We summarize the quantitative results on the number of parameters/FLOPs and the accuracy of several models/techniques to showcase the tradeoff exercised by them. We also outline future directions in this rapidly evolving field of research. We believe that this survey will educate both novice and seasoned researchers and also spark a plethora of research efforts in this field.

97 MATHEMATICS AND COMPUTING↗

Closure to “Observation-Based Evaluation of Flood Estimation Techniques for a Small Snow-Dominated Basin in the Washington Cascade Mountains”

Here, this paper presents a closure to “Observation-Based Evaluation of Flood Estimation Techniques for a Small Snow-Dominated Basin in the Washington Cascade Mountains” by Hongxiang Yan, Zhuoran Duan, Mark S. Wigmosta, Ning Sun, Ethan D. Gutmann, Bert Kruyt, and Jeffrey R. Arnold. https://doi.org/10.1061/JHYEFF .HEENG-6147.

54 ENVIRONMENTAL SCIENCES↗

ForQint

ForQint is a Fortran software library to adaptively integrate multivariable functions versus well known probability distributions (e.g. normal, lognormal, etc.) using quadrature rules. Such integration is known to be more efficient and accurate especially when the function evaluation is expensive. The library also FQint enables the creation of both full and sparse anisotropic multi-dimensional quadrature rules. These latter encompass different such as Gauss-Hermite, Gauss-Legendre, etc. The functions Maher Salloum, Bert Debusschere, Kenny Chowdhary and subroutines rely on the C++ Stokhos package of Trilinos (trilinos.sandia.gov) where the appropriate classes were wrapped into Fortran.

Chowdhary, Kamalijit↗

An Introduction to Word Embeddings and Language Models

Language models have advanced at a phenomenal pace over the past decade. This document provides a short introduction to terminology, word embeddings (aka low-dimensional representations), and popular large-scale language models (LMs). Word embeddings are used to represent words as numerical vectors and are context-independent, meaning a word can only have a single representation (e.g., club can only be club sandwich, not golf club ). Language models can determine the probability of a given sequence of words occurring in a sentence and can provide context to distinguish between words and phrases that sound similar. LMs are context-dependent (e.g., club can be club sandwich or golf club ) and largely fall in two main classes – autoregressive and autoencoding models. Autoregressive models are pretrained on the classic language modeling task: guess the next token having read all the previous ones. Those models can be fine-tuned and achieve great results on many tasks, the most natural application is text generation. A typical example of such models is GPT, but others include GPT-2, GPT-3, CTLR, TRANSFORMER-XL, REFORMER, XLNET. Autoencoding models are pretrained by corrupting the input tokens in some way and trying to reconstruct the original sentence. They can be fine-tuned and achieve great results on many tasks such as text generation, but their most natural application is sentence classification or token classification. A typical example of such models is BERT, but others include ROBERTA, ALBERT, XML, XML-ROBERTA, FLAUBERT AND LONGFORMER.

97 MATHEMATICS AND COMPUTING↗

2022 Spring Internship Exit Presentation

As efforts of the National Aeronautics and Space Administration (NASA) and the Federal Aviation Administration (FAA) continue to digitize the air traffic management (ATM) domain, there is countless times of need for downstream natural language processing (NLP) tasks such as named entity recognition, text summarization, classification, and more. Although there are a plethora of open-sourced pre-trained transformer models in the NLP field such as BERT, RoBERTa, XLNet, and GPT-3, these models are trained on general corpora and perform poorly on domain-specific terminology and phraseology seen in ATM documents such as Notice to Airmen (NOTAMs) and Letters of Agreement (LoA). Our proposed research objective will be to first gather a large corpus of air traffic management related documents, orders, notices, books, technical papers, conference papers, articles, and other miscellaneous sources of text data from the FAA, NASA, and accredited conference and publication societies. After gathering this data, many steps will have to be taken to collate and preprocess the data into a format understandable by our test transformer models. Thirdly, we will set up training pipelines to train the RoBERTa model on its unsupervised training task masked language modelling (MLM) using resources provided by the NASA Advanced Supercomputing (NAS) facilities. Finally, these fine-tuned transformer models will be evaluated on their performance on down-stream NLP tasks as mentioned above, to show whether they will be effective when working with ATM related data or not. Once complete, this model could be made open-sourced on the HuggingFace website, where the rest of the ATM community can access and utilize this tool.

NLP↗

Leveraging large language models to automate the identification of healthcare access barriers for veterans

Objective: To develop and evaluate an automated system for identifying healthcare barriers focusing on transportation issues in veterans’ clinical notes using large language models (LLMs) and to assess the impact of different prompting strategies on classification performance and explanation consistency. Methods: We developed a hybrid system combining pattern matching for templated notes with LLM analysis for free-text notes. Using 2000 manually annotated clinical notes, we compared four prompting strategies (dual-role short, dual-role long, analysis-first, analysis-only) across Mistral-7B and Llama-3.1 models. We evaluated classification performance using standard metrics and assessed explanation consistency through embedding similarity analysis. Results: The analysis-first strategy achieved superior performance, with Mistral-7B reaching an F1 score of 0.914, outperforming traditional machine learning approaches (GBM: 0.786, BERT: 0.811). LLMs demonstrated higher explanation consistency within models (mean cosine similarity 0.887–0.908) compared to cross-model similarities (0.767–0.872). Pattern matching successfully handled 6.7% of templated notes deterministically. Mistral-7B showed greater internal consistency but higher abstention rates compared to Llama-3.1. Conclusion: Requiring LLMs to analyze evidence before classification improves both accuracy and explanation consistency for identifying transportation barriers in clinical notes. This approach enables automated barrier detection at scale while providing clinically relevant explanations, supporting both population-level healthcare planning and individual patient care decisions.

Healthcare access barriers↗

Multiview Incomplete Knowledge Graph Integration with application to cross-institutional EHR data harmonization

Objective: The growing availability of electronic health records (EHR) data opens opportunities for integrative analysis of multi-institutional EHR to produce generalizable knowledge. A key barrier to such integrative analyses is the lack of semantic interoperability across different institutions due to coding differences. We propose a Multiview Incomplete Knowledge Graph Integration (MIKGI) algorithm to integrate information from multiple sources with partially overlapping EHR concept codes to enable translations between healthcare systems. Methods: The MIKGI algorithm combines knowledge graph information from (i) embeddings trained from the co-occurrence patterns of medical codes within each EHR system and (ii) semantic embeddings of the textual strings of all medical codes obtained from the Self-Aligning Pretrained BERT (SAPBERT) algorithm. Due to the heterogeneity in the coding across healthcare systems, each EHR source provides partial coverage of the available codes. MIKGI synthesizes the incomplete knowledge graphs derived from these multi-source embeddings by minimizing a spherical loss function that combines the pairwise directional similarities of embeddings computed from all available sources. MIKGI outputs harmonized semantic embedding vectors for all EHR codes, which improves the quality of the embeddings and enables direct assessment of both similarity and relatedness between any pair of codes from multiple healthcare systems. Results: With EHR co-occurrence data from Veteran Affairs (VA) healthcare and Mass General Brigham (MGB), MIKGI algorithm produces high quality embeddings for a variety of downstream tasks including detecting known similar or related entity pairs and mapping VA local codes to the relevant EHR codes used at MGB. Based on the cosine similarity of the MIKGI trained embeddings, the AUC was 0.918 for detecting similar entity pairs and 0.809 for detecting related pairs. For cross-institutional medical code mapping, the top 1 and top 5 accuracy were 91.0% and 97.5% when mapping medication codes at VA to RxNorm medication codes at MGB; 59.1% and 75.8% when mapping VA local laboratory codes to LOINC hierarchy. When trained with 500 labels, the lab code mapping attained top 1 and 5 accuracy at 77.7% and 87.9%. MIKGI also attained best performance in selecting VA local lab codes for desired laboratory tests and COVID-19 related features for COVID EHR studies. Compared to existing methods, MIKGI attained the most robust performance with accuracy the highest or near the highest across all tasks. Conclusions: The proposed MIKGI algorithm can effectively integrate incomplete summary data from biomedical text and EHR data to generate harmonized embeddings for EHR codes for knowledge graph modeling and cross-institutional translation of EHR codes.

Zhou, Doudou↗

A rule-free workflow for the automated generation of databases from scientific literature

Abstract In recent times, transformer networks have achieved state-of-the-art performance in a wide range of natural language processing tasks. Here we present a workflow based on the fine-tuning of BERT models for different downstream tasks, which results in the automated extraction of structured information from unstructured natural language in scientific literature. Contrary to existing methods for the automated extraction of structured compound-property relations from similar sources, our workflow does not rely on the definition of intricate grammar rules. Hence, it can be adapted to a new task without requiring extensive implementation efforts and knowledge. We test our data-extraction workflow by automatically generating a database for Curie temperatures and one for band gaps. These are then compared with manually curated datasets and with those obtained with a state-of-the-art rule-based method. Furthermore, in order to showcase the practical utility of the automatically extracted data in a material-design workflow, we employ them to construct machine-learning models to predict Curie temperatures and band gaps. In general, we find that, although more noisy, automatically extracted datasets can grow fast in volume and that such volume partially compensates for the inaccuracy in downstream tasks.

36 MATERIALS SCIENCE↗

CMS Full Simulation for Run 3

We report the status of the CMS full simulation for Run 3. During the long shutdown of the LHC a significant update has been introduced to the CMS code for simulation. The CMS geometry description is reviewed. Several important modifications were needed. CMS detector description software is migrated to the DD4Hep community developed tool. We will report on our experience obtained during the process of this migration. Geant4 10.7 is the CMS choice for Run 3 simulation productions. We will discuss arguments for this choice, the strategy of adaptation of a new Geant4 version, and will report on the physics performance of the CMS simulation. A special Geant4 Physics List configuration FTFP_BERT_EMM will be described, which provides a compromise between simulation accuracy and CPU performance. A significant fraction of time for simulation of CMS events is spent on tracking of charged particles in a magnetic field. In the CMS simulation a dynamic choice of Geant4 parameters for tracking in field is implemented. A new method is introduced into simulation of electromagnetic components of hadronic showers in the electromagnetic calorimeter of CMS. For low-energy electrons and positrons a parametrization of GFlash type is applied. Results of tests of this method will be discussed. In summary, we expect about 25% speedup of the CMS simulation production for Run 3 compared to the Run 2 simulations.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Full Simulation of CMS for Run-3 and Phase-2

In this contribution we report the status of the CMS Geant4 simulation and the prospects for Run-3 and Phase-2. Firstly, we report about our experience during the start of Run-3 with Geant4 10.7.2, the common software package DD4hep for geometry description, and VecGeom runtime geometry library. In addition, FTFP_BERT_EMM Physics List and CMS configuration for tracking in magnetic field have been utilized. For the first time, for the Grid mass production of Monte-Carlo, this combination of components is used. Further simulation improvements are under development targeting Run-3 such as the switch to the new Geant4 11.1 in production, that provides several features important for the optimization of simulation, for example the new transportation process with built-in multiple scattering, neutron general process, custom tracking manager, G4HepEm sub-library, and others. We will present evaluation of various options, validation results, and the final choice of simulation configuration for 2023 production and beyond. The performance of the CMS full simulation for Run-2 and Run-3 will also be discussed. CMS development plan for the Phase-2 Geant4 based simulation is very ambitious, and it includes a new geometry description, physics, and simulation configurations. The progress on new detector descriptions and full simulation will be presented as well as the R&D in progress to reduce compute capacity needs.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

K-RET: knowledgeable biomedical relation extraction system

Abstract Motivation Relation extraction (RE) is a crucial process to deal with the amount of text published daily, e.g. to find missing associations in a database. RE is a text mining task for which the state-of-the-art approaches use bidirectional encoders, namely, BERT. However, state-of-the-art performance may be limited by the lack of efficient external knowledge injection approaches, with a larger impact in the biomedical area given the widespread usage and high quality of biomedical ontologies. This knowledge can propel these systems forward by aiding them in predicting more explainable biomedical associations. With this in mind, we developed K-RET, a novel, knowledgeable biomedical RE system that, for the first time, injects knowledge by handling different types of associations, multiple sources and where to apply it, and multi-token entities. Results We tested K-RET on three independent and open-access corpora (DDI, BC5CDR, and PGR) using four biomedical ontologies handling different entities. K-RET improved state-of-the-art results by 2.68% on average, with the DDI Corpus yielding the most significant boost in performance, from 79.30% to 87.19% in F-measure, representing a P-value of 2.91×10−12. Availability and implementation https://github.com/lasigeBioTM/K-RET.

Sousa, Diana F. (ORCID:0000000305979273)↗

A Decision Support System to Compile Environmental Mitigations from Hydropower Licensing Documents

The process of deciphering, extracting, and compiling information from texts dense with domain-specific terminology and technical jargon is a challenging endeavor. It demands considerable expertise and deep knowledge in the respective field, resulting in a labor-intensive process when executed by humans. Furthermore, the task of identifying multiple class labels in extensive texts presents a challenge due to intra- and inter-reader variability, making the process time-consuming and costly.We’re introducing a user-friendly graphical interface, fortified with a BERT model-powered decision support system. This advanced system aims to augment efficiency, curtail data collection time, and sustain high precision in data acquisition. It is instrumental in deciphering and synthesizing intricate texts teeming with a spectrum of expressions, even within similar mitigation categories. Such tasks traditionally demand substantial human effort and specialized knowledge in the domain.Our system is specifically engineered for the task of extracting environmental mitigation information to promote sustainable hydropower development from licenses issued by the Federal Energy Regulatory Commission (FERC). These license documents are comprehensive, each containing over 15,000 words and requiring the identification of 135 different class labels. We anticipate that our system will boost reading speed, improve the consistency of classification outputs among readers, and contribute to the development of a robust scientific database of environmental mitigations associated with the 2,000+ non-federal hydropower facilities licensed by FERC in the United States.

Yoon, Hong-Jun [ORNL] (ORCID:0000000254505878)↗

Efficient Clustering of Software Vulnerabilities using Self Organizing Map (SOM)

The common vulnerabilities and exposures (CVE) database was created with a mission to ``identify, define, and catalog publicly disclosed cybersecurity vulnerabilities''. This rich body of information can be used to enable rapid and efficient response to secure and defend cyber operations and protect critical cyber infrastructure. The main goal of this paper is to develop a visual analytics tool to enable deep analysis of CVEs using unsupervised clustering techniques. We enhance our analysis by first mapping CVEs to hierarchical-classes in Common Weakness Enumeration (CWE) using information in the National Vulnerability Database (NVD). Both the mapping and the numerical representation of CVEs are enabled by V2W-BERT, which uses natural language processing of the extensive information in NVD to generate a large tabular database of 137,226 CVE entries from 1999 to 2020, where each CVE is represented by a vector of 768 numerical features. The vectorized data is processed by Self-Organizing Maps (SOM), which is an unsupervised machine learning technique for dimensionality reduction, visual representation and clustering. Using a Torus map of 6417 units, we achieve ~10-fold data compression of ~140k CVEs using SOM. The trained map is further clustered using standard K-means clustering into 138 clusters of CVEs. We conducted a brief investigation of the rich mapping of CVEs to best-matching-units to K-means clusters, as well as CVEs to CWEs. For example, this novel mapping provided insight into the role of CWE-59 and CWE-264 in several CVEs that is otherwise hard to explore in the original data. We conclude that our this novel approach will not only enable deep analysis of the complex relationships between CVEs and CWEs, but also a mechanism to quickly respond to and design mitigation actions for rapidly evolving vulnerabilities that have not been mapped to existing CWEs.

Panchal, Khyati↗

Energy–Performance Trade-offs in Privacy-Preserving Federated Learning on SmartNIC-Enabled HPC Systems

Federated learning (FL) is increasingly deployed on accelerator-rich high-performance computing (HPC) systems, yet the system-level energy cost of privacy-aware FL remains poorly understood, particularly across heterogeneous networking and server-placement options. We present a measurement-driven study of energy–performance trade-offs for FL on GH200-class nodes across three deployment configurations: CPU-Ethernet, CPU-InfiniBand (RDMA-capable), and a DPU-hosted FL server over InfiniBand using a BlueField-3 SmartNIC/DPU. Using NVIDIA FLARE (NVFLARE), we align node-level power telemetry with per-round timing extracted from NVFLARE logs to quantify time-to-solution (TTS), energy-to-solution (ETS), energy-delay product (EDP), and synchronization behavior for three transformer models (ALBERT, DistilBERT, BERT), trained with and without differential privacy (DP). We find that interconnect choice is the dominant driver of runtime and energy: host-managed InfiniBand consistently reduces communication overhead versus Ethernet, yielding lower TTS/ETS/EDP. In contrast, in our NVFLARE deployment, placing the FL server on the DPU does not consistently match CPU-InfiniBand performance and can be slower—especially for larger models—highlighting that server placement alone is not sufficient to guarantee end-to-end gains. Finally, under our fixed-round protocol, DP increases per-round cost and runtime variance; ETS increases largely in proportion to TTS because average node power remains relatively stable across configurations.

Kotevska, Olivera [ORNL] (ORCID:0000000316772243)↗

A Large Language Model for Determining Partial Tripping of Distributed Energy Resources

Knowing the status of individual distributed energy resources, i.e., being tripped or not, after a contingency can inform the development of an aggregated DER model. Here, this letter presents a large language model application to determine the partial tripping of distributed energy resources depending on the types, locations, and duration of faults in the transmission network. The large language model, or more specifically BERT-based approach can streamline the fault information into tokenized input, which not only reduces the complexity of the machine learning model but also demonstrates a robust performance with only limited data sets.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗