SEARCH · Search NASA
Results for “text 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.
Data-driven analysis of text-mined seed-mediated syntheses of gold nanoparticles
Gold nanoparticle synthesis recipes were extracted from the literature to obtain data-driven hypotheses for synthesis outcome morphology and size. Used images from https://Flaticon.com.
Trend Analysis in Biomedical Texts via Vector Space Model Synonym Extraction
Accurate synonym extraction is a highly desired tool in computational semantics and natural language processing (NLP). Intrinsic to many current language models is the Vector Space Model (VSM) whereby linguistic properties are modeled by real vectors of small length compared to the document size. Here, several vector models are discussed. Utilizing the open source Word2Vec tool, temporal trends in biomedical abstracts from MEDLINE have been analyzed via a unique transformation matrix method. Additionally, accuracy of the word vector method and clustering of temporal trends are discussed.
Text Stream Trend Analysis using Multiscale Visual Analytics with Applications to Social Media Systems
In this paper, we introduce a new visual analytics system, called Matisse, that allows exploration of global trends in textual information streams with specific application to social media platforms. Despite the potential for real-time situational awareness using these services, interactive analysis of such semi-structured textual information is a challenge due to the high-throughput and high-velocity properties. Matisse addresses these challenges through the following contributions: (1) robust stream data management, (2) automated sen- timent/emotion analytics, (3) inferential temporal, geospatial, and term-frequency visualizations, and (4) a flexible drill-down interaction scheme that progresses from macroscale to microscale views. In addition to describing these contributions, our work-in-progress paper concludes with a practical case study focused on the analysis of Twitter 1% sample stream information captured during the week of the Boston Marathon bombings.
Identification of candidate genes in Arabidopsis and Populus cell wall biosynthesis using text-mining, co-expression network analysis and comparative genomics
Explore the source record for details and available documents.
NASA's online machine aided indexing system
This report describes the NASA Lexical Dictionary, a machine aided indexing system used online at the National Aeronautics and Space Administration's Center for Aerospace Information (CASI). This system is comprised of a text processor that is based on the computational, non-syntactic analysis of input text, and an extensive 'knowledge base' that serves to recognize and translate text-extracted concepts. The structure and function of the various NLD system components are described in detail. Methods used for the development of the knowledge base are discussed. Particular attention is given to a statistically-based text analysis program that provides the knowledge base developer with a list of concept-specific phrases extracted from large textual corpora. Production and quality benefits resulting from the integration of machine aided indexing at CASI are discussed along with a number of secondary applications of NLD-derived systems including on-line spell checking and machine aided lexicography.
Language Model For Earth Science: Exploring Potential Downstream Applications As Well As Current Challenges
The use of deep learning techniques to build transformer language models such as SciBERT and GPT3 have transformed the natural language technology (NLT) landscape. These new NLTs are being used in speech to text and vice versa, auto-mated text classification, sentiment analysis, topic modeling, text summarization, and cognitive assistants. While Earth science has no shortage of unstructured data such as journal and conference papers, little efforts have focused on harnessing NLTs for knowledge extraction and supporting the scientific process. This paper surveys the use of language models in different science. BERT-E, a new Earth science-specific language model, is presented. BERT-E is generated using a transfer learning solution. A language model that has already been trained for general Science (SciBERT) is fine-tuned using abstracts and full text extracted from various Earth science-related articles. A downstream keywords classification application is used for evaluation, and the use of BERT-E shows improved performance. The need to develop a robust set of benchmarks in evaluating the language model such as BERT-E is discussed. Finally, example applications are presented to inspire additional ideas for applications using domain-specific language models.
Interactive Visual Analysis of High Throughput Text Streams
The scale, velocity, and dynamic nature of large scale social media systems like Twitter demand a new set of visual analytics techniques that support near real-time situational awareness. Social media systems are credited with escalating social protest during recent large scale riots. Virtual communities form rapidly in these online systems, and they occasionally foster violence and unrest which is conveyed in the users language. Techniques for analyzing broad trends over these networks or reconstructing conversations within small groups have been demonstrated in recent years, but state-of- the-art tools are inadequate at supporting near real-time analysis of these high throughput streams of unstructured information. In this paper, we present an adaptive system to discover and interactively explore these virtual networks, as well as detect sentiment, highlight change, and discover spatio- temporal patterns.
Text Mining Improves Prediction of Protein Functional Sites
We present an approach that integrates protein structure analysis and text mining for protein functional site prediction, called LEAP-FS (Literature Enhanced Automated Prediction of Functional Sites). The structure analysis was carried out using Dynamics Perturbation Analysis (DPA), which predicts functional sites at control points where interactions greatly perturb protein vibrations. The text mining extracts mentions of residues in the literature, and predicts that residues mentioned are functionally important. We assessed the significance of each of these methods by analyzing their performance in finding known functional sites (specifically, small-molecule binding sites and catalytic sites) in about 100,000 publicly available protein structures. The DPA predictions recapitulated many of the functional site annotations and preferentially recovered binding sites annotated as biologically relevant vs. those annotated as potentially spurious. The text-based predictions were also substantially supported by the functional site annotations: compared to other residues, residues mentioned in text were roughly six times more likely to be found in a functional site. The overlap of predictions with annotations improved when the text-based and structure-based methods agreed. Our analysis also yielded new high-quality predictions of many functional site residues that were not catalogued in the curated data sources we inspected. We conclude that both DPA and text mining independently provide valuable high-throughput protein functional site predictions, and that integrating the two methods using LEAP-FS further improves the quality of these predictions.
SU-C-BRD-03: Analysis of Accelerator Generated Text Logs for Preemptive Maintenance
Purpose: To develop a model to analyze medical accelerator generated parameter and performance data that will provide an early warning of performance degradation and impending component failure. Methods: A robust 6 MV VMAT quality assurance treatment delivery was used to test the constancy of accelerator performance. The generated text log files were decoded and analyzed using statistical process control (SPC) methodology. The text file data is a single snapshot of energy specific and overall systems parameters. A total of 36 system parameters were monitored which include RF generation, electron gun control, energy control, beam uniformity control, DC voltage generation, and cooling systems. The parameters were analyzed using Individual and Moving Range (I/MR) charts. The chart limits were calculated using a hybrid technique that included the use of the standard 3σ limits and the parameter/system specification. Synthetic errors/changes were introduced to determine the initial effectiveness of I/MR charts in detecting relevant changes in operating parameters. The magnitude of the synthetic errors/changes was based on: the value of 1 standard deviation from the mean operating parameter of 483 TB systems, a small fraction (≤ 5%) of the operating range, or a fraction of the minor fault deviation. Results: There were 34 parameters in which synthetic errors were introduced. There were 2 parameters (radial position steering coil, and positive 24V DC) in which the errors did not exceed the limit of the I/MR chart. The I chart limit was exceeded for all of the remaining parameters (94.2%). The MR chart limit was exceeded in 29 of the 32 parameters (85.3%) in which the I chart limit was exceeded. Conclusion: Statistical process control I/MR evaluation of text log file parameters may be effective in providing an early warning of performance degradation or component failure for digital medical accelerator systems. Research is Supported by Varian Medical Systems, Inc.
Measurement of prompt and nonprompt charmonium suppression in $\text {PbPb}$ collisions at 5.02 $\,\text {Te}\text {V}$
The nuclear modification factors of ${\mathrm {J}/\psi }$ and $\psi \text {(2S)}$ mesons are measured in $\text {PbPb}$ collisions at a centre-of-mass energy per nucleon pair of $\sqrt{\smash [b]{s_{_{\text {NN}}}}} = 5.02\,\text {Te}\text {V} $ . The analysis is based on $\text {PbPb}$ and $\mathrm {p}\mathrm {p}$ data samples collected by CMS at the LHC in 2015, corresponding to integrated luminosities of 464 $\,\mu \mathrm {b}^{-1}$ and 28 $\,\text {pb}^\text {-1}$ , respectively. The measurements are performed in the dimuon rapidity range of $|y | < 2.4$ as a function of centrality, rapidity, and transverse momentum ( $p_{\mathrm {T}}$ ) from $p_{\mathrm {T}} =3$ ${\,\text {Ge}\text {V}/}\text {c}$ in the most forward region and up to 50 ${\,\text {Ge}\text {V}/}\text {c}$ . Both prompt and nonprompt (coming from b hadron decays) ${\mathrm {J}/\psi }$ mesons are observed to be increasingly suppressed with centrality, with a magnitude similar to the one observed at $\sqrt{\smash [b]{s_{_{\text {NN}}}}} = 2.76\,\text {Te}\text {V} $ for the two ${\mathrm {J}/\psi }$ meson components. No dependence on rapidity is observed for either prompt or nonprompt ${\mathrm {J}/\psi }$ mesons. An indication of a lower prompt ${\mathrm {J}/\psi }$ meson suppression at $p_{\mathrm {T}} > 25$ ${\,\text {Ge}\text {V}/}\text {c}$ is seen with respect to that observed at intermediate $p_{\mathrm {T}}$ . The prompt $\psi \text {(2S)}$ meson yield is found to be more suppressed than that of the prompt ${\mathrm {J}/\psi }$ mesons in the entire $p_{\mathrm {T}}$ range.
A Scalable Visual System for Proliferation Analysis Phase I Final Report
Explore the source record for details and available documents.
Comprehension and retrieval of failure cases in airborne observatories
This paper describes research dealing with the computational problem of analyzing and repairing failures of electronic and mechanical systems of telescopes in NASA's airborne observatories, such as KAO (Kuiper Airborne Observatory) and SOFIA (Stratospheric Observatory for Infrared Astronomy). The research has resulted in the development of an experimental system that acquires knowledge of failure analysis from input text, and answers questions regarding failure detection and correction. The system's design builds upon previous work on text comprehension and question answering, including: knowledge representation for conceptual analysis of failure descriptions, strategies for mapping natural language into conceptual representations, case-based reasoning strategies for memory organization and indexing, and strategies for memory search and retrieval. These techniques have been combined into a model that accounts for: (a) how to build a knowledge base of system failures and repair procedures from descriptions that appear in telescope-operators' logbooks and FMEA (failure modes and effects analysis) manuals; and (b) how to use that knowledge base to search and retrieve answers to questions about causes and effects of failures, as well as diagnosis and repair procedures. This model has been implemented in FANSYS (Failure ANalysis SYStem), a prototype text comprehension and question answering program for failure analysis.
Lexicon generation methods, lexicon generation devices, and lexicon generation articles of manufacture
Lexicon generation methods, computer implemented lexicon editing methods, lexicon generation devices, lexicon editors, and articles of manufacture are described according to some aspects. In one aspect, a lexicon generation method includes providing a seed vector indicative of occurrences of a plurality of seed terms within a plurality of text items, providing a plurality of content vectors indicative of occurrences of respective ones of a plurality of content terms within the text items, comparing individual ones of the content vectors with respect to the seed vector, and responsive to the comparing, selecting at least one of the content terms as a term of a lexicon usable in sentiment analysis of text.
NOUS: A Knowledge Graph Management System
Knowledge graphs represent information as entities and relationships between them. For tasks such as natural language question answering or automated analysis of text, a knowledge graph provides valuable context to establish the specific type of entities being discussed. It allow us to derive better context about newly arriving information and leads to intelligent reasoning capabilities. We address two primary needs: A) Automated construction of knowledge graphs is a technically challenging, expensive process; and B) The ability to synthesize new information by monitoring newly emerging knowledge is a transformational capability that does not exist in state of the art systems.
Python HITMIX
SAND2024-10158O Python HITMIX is a software package for computing hitting time moments of vertices in a graph. Hitting time moments can be used to rank the strengths of relationships between vertices in a graph. Examples are provided for computing and using hitting time moments on generic graphs and similarity graphs arising from the analysis of text documents in information retrieval applications. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.
Information Theoretic Measures for Visual Analytics: The Silver Ticket?
In the context of text-based analysis workflows, we propose that an effective analytic tool facilitates triage by a) enabling users to identify and set aside irrelevant content (i.e., reduce the complexity of information in a dataset) and b) develop a working mental model of which items are most relevant to the question at hand. This LDRD funded research developed a dataset that is enabling this team to evaluate propose normalized compression distance (NCD) as a task, user, and context-insensitive measure of categorization outcomes (Shannon entropy is reduced as order is imposed). Effective analytics tools help people impose order, reducing complexity in measurable ways. Our concept and research was documented in a paper accepted to the ACM conference "Beyond Time and Error: Novel Methods in Information Visualization Evaluation", part of the IEEE VisWeek Conference, Baltimore, MD, October 16-21, 2016. The paper is included as an appendix to this report.
Morphing Terminology Study
This study investigates methods of automatically identifying and characterizing significant transitions in term usage over time. Within scientific literature, the occurrence of terms reflects the use of technologies and techniques as well as the study of specific species and materials. Transitions in terminology usage may be a result of vocabulary standardization or specialization in which terms are replaced with their shorter form. They may also be a result of new applications, combinations, alternatives, or interests that result in the appearance of new or existing terminology in unexpected contexts.