Natural Language Processing Text Research Articles

Challenges of Developing a Natural Language Processing Method With Electronic Health Records to Identify Persons With Chronic Mobility Disability

ObjectiveTo assess the utility of applying natural language processing (NLP) to electronic health records (EHRs) to identify individuals with chronic mobility disability. DesignWe used EHRs from the Research Patient Data Repository, which contains EHRs from a large Massachusetts health care delivery system. This analysis was part of a larger study assessing the effects of disability on diagnosis of colorectal cancer. We applied NLP text extraction software to longitudinal EHRs of colorectal cancer patients to identify persons who use a wheelchair (our indicator of mobility disability for this analysis). We manually reviewed the clinical notes identified by NLP using directed content analysis to identify true cases using wheelchairs, duration or chronicity of use, and documentation quality. SettingEHRs from large health care delivery system ParticipantsPatients (N=14,877) 21-75 years old who were newly diagnosed with colorectal cancer between 2005 and 2017. InterventionsNot applicable. Main Outcome MeasuresConfirmation of patients’ chronic wheelchair use in NLP-flagged notes; quality of disability documentation. ResultsWe identified 14,877 patients with colorectal cancer with 303,182 associated clinical notes. NLP screening identified 1482 (0.5%) notes that contained 1+ wheelchair-associated keyword. These notes were associated with 420 patients (2.8% of colorectal cancer population). Of the 1482 notes, 286 (19.3%, representing 105 patients, 0.7% of the total) contained documentation of reason for wheelchair use and duration. Directed content analysis identified 3 themes concerning disability documentation: (1) wheelchair keywords used in specific EHR contexts; (2) reason for wheelchair not clearly stated; and (3) duration of wheelchair use not consistently documented. ConclusionsNLP offers an option to screen for patients with chronic mobility disability in much less time than required by manual chart review. Nonetheless, manual chart review must confirm that flagged patients have chronic mobility disability (are not false positives). Notes, however, often have inadequate disability documentation.

Generation and evaluation of artificial mental health records for Natural Language Processing

A serious obstacle to the development of Natural Language Processing (NLP) methods in the clinical domain is the accessibility of textual data. The mental health domain is particularly challenging, partly because clinical documentation relies heavily on free text that is difficult to de-identify completely. This problem could be tackled by using artificial medical data. In this work, we present an approach to generate artificial clinical documents. We apply this approach to discharge summaries from a large mental healthcare provider and discharge summaries from an intensive care unit. We perform an extensive intrinsic evaluation where we (1) apply several measures of text preservation; (2) measure how much the model memorises training data; and (3) estimate clinical validity of the generated text based on a human evaluation task. Furthermore, we perform an extrinsic evaluation by studying the impact of using artificial text in a downstream NLP text classification task. We found that using this artificial data as training data can lead to classification results that are comparable to the original results. Additionally, using only a small amount of information from the original data to condition the generation of the artificial data is successful, which holds promise for reducing the risk of these artificial data retaining rare information from the original data. This is an important finding for our long-term goal of being able to generate artificial clinical data that can be released to the wider research community and accelerate advances in developing computational methods that use healthcare data.

«ДЕПУТАТСКАЯ КАРТОГРАФИЯ» НА МАТЕРИАЛЕ СТЕНОГРАММ ЗАСЕДАНИЙ ГД РФ

Public speeches of the deputies of the Russian Federation State Duma represent productive empirical material that requires an interdisciplinary approach to its analysis. The goal of the reported study is defining the degree of the concord regarding the future of Russia among deputies and the localization of this agreement in the semantic space (equilibrium point). The empirical base under consideration comprises the transcripts of Russian Federation State Duma sittings. This dataset covers time period from 1994 to mid-2020 and includes 324 thousand phrases (27 million words) from 2773 deputies and other people. The paper presents the general design of the study of the State Duma deputies discourse using the "Semograph" information system, the SlovNet library for natural language text processing based on deep learning, the SciVi visual analytics platform, the map visualization module based on the Leaflet library, the geocoding of geographic objects based on the OpenStreetMap map provider and other tools. The application of the approach showed the importance of the institutional foundations of the symbolic deputies' mapping of the regions of the Volga Federal District (the electoral connection of a deputy with the region – deputies elected by party list or pluralist rule).

Attention-based Sentiment Reasoner for aspect-based sentiment analysis

Aspect-based sentiment analysis (ABSA) is a powerful way of predicting the sentiment polarity of text in natural language processing. However, understanding human emotions and reasoning from text like a human continues to be a challenge. In this paper, we propose a model, named Attention-based Sentiment Reasoner (AS-Reasoner), to alleviate the problem of how to capture precise sentiment expressions in ABSA for reasoning. AS-Reasoner assigns importance degrees to different words in a sentence to capture key sentiment expressions towards a specific aspect, and transfers them into a sentiment sentence representation for reasoning in the next layer. To obtain appropriate importance degree values for different words in a sentence, two attention mechanisms we designed: intra attention and global attention. Specifically, intra attention captures the sentiment similarity between any two words in a sentence to compute weights and global attention computes weights by a global perspective. Experiments on all four English and four Chinese datasets show that the proposed model achieves state-of-the-art accuracy and macro-F1 results for aspect term level sentiment analysis and obtains the best accuracy for aspect category level sentiment analysis. The experimental results also indicate that AS-Reasoner is language-independent.

An Erroneous Word Locating Algorithm for Natural Language Text in Medical Scene Based on Clustering Analysis

With the development of technology, the processing of natural language texts is increasingly used in medical scenes. For example, speech recognition technology and smart customer service robots can help hospitals to improve efficiency. However, the accuracy rate of the text output of the speech recognition is not satisfactory, especially in the medical scene. As a result, the study of natural language text processing in medical scene is very meaningful. To address this issue, we construct a corpus, a knowledge base, a Pinyin library and a word frequency library. We perform word segmentation operations based on the knowledge base and calculate the word co-occurrence probability based on the word frequency library. The word co-occurrence probability is used to determine whether erroneous words exist or not. If there are words whose co-occurrence probabilities are below a threshold, we assert that there exists erroneous words. And then the clustering analysis is used to locate the erroneous words. The experimental results show that the accuracy rate of erroneous word locating is very high. Therefore, this method has certain practical value.

Log mining and knowledge-based models in data storage systems diagnostics

Modern data storage systems have a sophisticated hardware and software architecture, including multiple storage processors, storage fabrics, network equipment and storage media and contain information, which can be damaged or lost because of hardware or software fault. Approach to storage software diagnostics, presented in current paper, combines a log mining algorithms for fault detection based on natural language processing text classification methods, and usage of the diagnostic model for a task of fault source detection. Currently existing approaches to computational systems diagnostics are either ignoring system or event log data, using only numeric monitoring parameters, or target only certain log types or use logs to create chains of the structured events. The main advantage of using natural language processing method for log text classification is that no information of log message structure or log message source, or log purpose is required if there is enough data for classificator model training. Developed diagnostic procedure has accuracy score comparable with existing methods and can target all presented in training set faults without prior log structure research.

A Named Entity Recognition System for Malayalam using Neural Networks

Named Entity Recognition is the process of identifying the entities in the text document and categorizing them into predefined categories such as Person, Location, Organisation, etc. It is an important step in the processing of natural language text. Named entity recognition systems aim at extracting relevant information from the text. Various methods are applied for NER in Malayalam. In this paper, we propose an NER system for Malayalam using neural networks. Neural networks are exceptionally powerful tools for learning representations of data with multiple levels of abstraction. The proposed system utilizes different features such as POS information of the word, embedded representation of words and suffixes, POS information of preceding words, etc. We have used a corpus of 20615 sentences for training and testing. With less number of features, the system was able to obtain the state of the art performance in NER for Malayalam.

Toward high-throughput phenotyping: unbiased automated feature extraction and selection from knowledge sources

Analysis of narrative (text) data from electronic health records (EHRs) can improve population-scale phenotyping for clinical and genetic research. Currently, selection of text features for phenotyping algorithms is slow and laborious, requiring extensive and iterative involvement by domain experts. This paper introduces a method to develop phenotyping algorithms in an unbiased manner by automatically extracting and selecting informative features, which can be comparable to expert-curated ones in classification accuracy. Comprehensive medical concepts were collected from publicly available knowledge sources in an automated, unbiased fashion. Natural language processing (NLP) revealed the occurrence patterns of these concepts in EHR narrative notes, which enabled selection of informative features for phenotype classification. When combined with additional codified features, a penalized logistic regression model was trained to classify the target phenotype. The authors applied our method to develop algorithms to identify patients with rheumatoid arthritis and coronary artery disease cases among those with rheumatoid arthritis from a large multi-institutional EHR. The area under the receiver operating characteristic curves (AUC) for classifying RA and CAD using models trained with automated features were 0.951 and 0.929, respectively, compared to the AUCs of 0.938 and 0.929 by models trained with expert-curated features. Models trained with NLP text features selected through an unbiased, automated procedure achieved comparable or slightly higher accuracy than those trained with expert-curated features. The majority of the selected model features were interpretable. The proposed automated feature extraction method, generating highly accurate phenotyping algorithms with improved efficiency, is a significant step toward high-throughput phenotyping.

An Experimental Study of Incremental SVD on Latent Semantic Analysis

Singular value decomposition is an important method to factorize a matrix, which is useful for many applications, such as principal component analysis in statistics and latent semantic analysis in natural language text processing. In order to diminish the computational complexity, incremental update algorithm is developed, especially for analyzing relationships between text documents. However, it is at the cost of a reduction in precision. In this paper, we study the effect of incremental update on latent semantic analysis; and, by experiments, to investigate the trend downward of computing precision and then try to find the balance point between re-computation and incremental update. The results indicate that when increased documents are less than original documents, incremental update method has an approximate precision to re-computing method, and then its recall ratio will obviously decline. It could be helpful for deciding when the matrix factorization must be recomputed.

Classifying free-text triage chief complaints into syndromic categories with natural language processing

Objective: Develop and evaluate a natural language processing application for classifying chief complaints into syndromic categories for syndromic surveillance. Introduction: Much of the input data for artificial intelligence applications in the medical field are free-text patient medical records, including dictated medical reports and triage chief complaints. To be useful for automated systems, the free-text must be translated into encoded form. Methods: We implemented a biosurveillance detection system from Pennsylvania to monitor the 2002 Winter Olympic Games. Because input data was in free-text format, we used a natural language processing text classifier to automatically classify free-text triage chief complaints into syndromic categories used by the biosurveillance system. The classifier was trained on 4700 chief complaints from Pennsylvania. We evaluated the ability of the classifier to classify free-text chief complaints into syndromic categories with a test set of 800 chief complaints from Utah. Results: The classifier produced the following areas under the ROC curve: Constitutional = 0.95; Gastrointestinal = 0.97; Hemorrhagic = 0.99; Neurological = 0.96; Rash = 1.0; Respiratory = 0.99; Other = 0.96. Using information stored in the system’s semantic model, we extracted from the Respiratory classifications lower respiratory complaints and lower respiratory complaints with fever with a precision of 0.97 and 0.96, respectively. Conclusion: Results suggest that a trainable natural language processing text classifier can accurately extract data from free-text chief complaints for biosurveillance.

Natural language processing

Beginning with the basic issues of NLP, this chapter aims to chart the major research activities in this area since the last ARIST Chapter in 1996 (Haas, 1996), including: (i) natural language text processing systems - text summarization, information extraction, information retrieval, etc., including domain-specific applications; (ii) natural language interfaces; (iii) NLP in the context of www and digital libraries ; and (iv) evaluation of NLP systems.

A Knowledge-Based Approach to Conceptual Information Retrieval from Full Text

The conventional methods of full text processing, such as pattern matching and keyword indexing, may result in waste of rich information. Recently, the knowledge-based approach has been used in natural language text processing. Unlike past generations of information retrieval systems which dealt exclusively with abstracts, title headings, and keywords, today’s intelligent information retrieval systems manipulate full texts such as technical manuals, reports, letters, and news articles. This paper discusses the design and implementation of a knowledge-based information retrieval system, concentrating on the issues of knowledge acquisition, knowledge representation, knowledge base refinement, and the information retrieval mechanism

SCISOR: extracting information from on-line news

The future of natural language text processing is examined in the SCISOR prototype. Drawing on artificial intelligence techniques, and applying them to financial news items, this powerful tool illustrates some of the future benefits of natural language analysis through a combination of bottom-up and top-down processing.

Automated analysis of instructional text

The development of a capability for automated processing of natural language text is a long-range goal of artificial intelligence. This paper discusses an investigation into the issues involved in the comprehension of descriptive, as opposed to illustrative, textual material. The comprehension process is viewed as the conversion of knowledge from one representation into another. The proposed target representation consists of statements of the prolog language, which can be interpreted both declaratively and procedurally, much like production rules. A computer program has been written to model in detail some ideas about this process. The program successfully analyzes several heavily edited paragraphs adapted from an elementary textbook on programming, automatically synthesizing as a result of the analysis a working prolog program which, when executed, can parse and interpret LET commands in the basic language. The paper discusses the motivations and philosophy of the project, the many kinds of prerequisite knowledge which are necessary, and the structure of the text analysis program. A sentence-by-sentence account of the analysis of the sample text is presented, describing the syntactic and semantic processing which is involved. The paper closes with a discussion of lessons learned from the project, possible alternative approaches, and possible extensions for future work. The entire project is presented as illustrative of the nature and complexity of the text analysis process, rather than as providing definitive or optimal solutions to any aspects of the task.

Toward computational description of written Polish

We present briefly the activity of the Warsaw University. computational linguistics team. The team is an informal group that consolidates around a seminar at the Institute of Informatics, coordinated by Prof. Zygmunt Saloni. The group adheres to a "bootstrapping methodology": we focus on developing surface linguistic descriptions which should allow the creation of efficient tools for low-level processing of natural language texts, in the hope that it will pay when more sophisticated problems are attacked.

The EPISTLE project

The experimental EPISTLE system is ultimately intended to provide intelligent applications for the processing of natural language text. Potential applications are of two types: (1) processing of alreadygenerated text, and (2) assisting in the generation of new text.

COCOA: A FORTRAN program for concordance and word-count processing of natural language texts

Description. The COCOA program is a highly flexible device for processing natural language texts. The program can produce word and vocabulary counts, concordances with context formats and location references, and word-frequency profiles. Texts supplied by the user may be in any language, although non-Roman characters or alphabets must be transliterated by means of characters or combinations available on input devices and acceptable to the computer on which the program is to be run. Texts may be analyzed with a referencing that makes possible the separate processing of textual or library subdivisions (e.g., author, title, chapter, speaker, page, etc.). Concordances may be produced with various lengths and formats of line contexts for each vocabulary word, keywords, and keyword groups. Word counts may be arranged in alphabetical and reverse alphabetical order, frequency order, and counts restricted by declared frequency. To make the program adaptable to diverse research requirements, the individual user may declare his own: (a) alphabet, punctuation marks, word separators, and word connectors; (b) multiple character letters for transliteration of non-Roman alphabets; (c) special signs to distinguish homographs in program processing and printouts; (d) special characters for denoting word groups to be processed together; (e) references to particular subdivisions of a text or text library for sub text processing; (f) reference format for concordances; (g) declaration for listing word co-occurences within a desired text proximity; (h) inclusion or exclusion of a specified word or words; and (i) list of specified suffixes declared for special processing by the user. Input. Three files are required to operate the COCOA program: (1) the all-FORTRAN IV COCOA program, (2) a 13-card which includes alphabet and special declarations, word length, text-selection statements, and concordance or word-count requests; and (3) the language text. The version of COCOA adapted to the DECsystem-10 computer requires only two input streams. The control file and text file are collapsed, and may be input with a single card deck, and retained on user disk files or DEC-Tape. Experience on the DECsystem-10 indicates maximum efficiency is attained by punching the language text onto IBM tab cards, creating a disk file of the text, using a standard editing routine on a time-sharing terminal to correct typographical errors, and then storing each text on tape for recall when the particular test is required for processing. The COCOA program (521 blocks) may be stored on the file structure or user disk file. Operating efficiency is greatly increased when the COCOA program is stored in a compiled, machine-readable form for immediate execution. Prior to program execution, three input/output files and four work files must be assigned. When the user executes the COCOA program, the reads the control file-from which it receives character, text, and format instructions, then it reads and processes the text in accordance with the control requests. The data which is thus generated is read onto two output disk files. Output. The output file contains the system diagnostics, including a copy of the control file and any COCOA error messages generated during an unsuccessful or incomplete program execution. The other output file is a record of the word counts, word and vocabulary totals, frequency profile, and text concordance. These tables may be printed from the user disk files by a simple lineprinter command. Restrictions. COCOA handles alphabets of up to 256 letters of one per letter, 128 letters at two characters, and 85 letters at three characters. The length of texts is theoretically infinite, and depends only upon the available core space (28K) and the user's disk space quota. A 2,000-word text may be processed with a log-in quota of 1,000 blocks.