Large Text Corpora Research Articles

Comprehension and production of Kinyarwanda verbs in the Discriminative Lexicon

Abstract The Discriminative Lexicon is a theory of the mental lexicon that brings together insights from various other theories: words are the relevant cognitive units in morphology, the meaning of a word is represented by its distribution in utterances, word forms and their meaning are learned by minimizing prediction errors, and fully connected networks successfully capture language learning. In this article we model comprehension and production of Kinyarwanda verb forms in the Discriminative Lexicon model. Kinyarwanda is a highly inflectional language, and therefore particularly interesting, because its paradigms are almost unlimited in size. Can knowledge of its enormous paradigms be modeled only on the basis of words? To answer this question we modeled a data set of 11,528 verb forms, hand-annotated for meaning and their grammatical functions, in the Linear Discriminative Learning (LDL), a two-layered, fully connected computational implementation of the Discriminative Lexicon model. We also extracted 573 verbs from our data set for which meanings are available that are based on empirical word embeddings obtained from large text corpora, and modeled them in LDL. Both comprehension and production is learned accurately: Kinyarwanda verb forms can be comprehended and produced relying on words as cognitive units, in a two-layered network, in which prediction errors are minimized.

Media Representations of Healthcare Robotics in Norway 2000-2020: A Topic Modeling Approach

Robots are projected to affect healthcare services in significant, but unpredictable, ways. Many believe robots will add value to future healthcare, but their arrival has triggered controversy. Debates revolve around how robotics will impact healthcare provision, their effects on the future of labor and caregiver–patient relationships, and ethical dilemmas associated with autonomous machines. This study investigates media representations of healthcare robotics in Norway over a twenty-year period, using a mixed-methods design. Media representations affect public opinion in multiple ways. By assembling and presenting information through stories, they not only set the agenda by broadcasting values, experiences, and expectations about new technologies, but also frame and prime specific understandings of issues. First, we employ an inductive text-mining approach known as “topic modeling,” a computational method for eliciting abstract semantic structures from large text corpora. Using Non-Negative Matrix Factorization, we implement a topic model of manifest content from 752 articles, published in Norwegian print media between 1.1.2000 and 2.10.2020, sampled from a comprehensive database for news media (Atekst, Retriever). We complement this computational lens with a more fine-grained, qualitative analysis of content in exemplary texts sampled from each topic. Here, we identify prominent “frames,” discursive cues for interpreting how various stakeholders talk about healthcare robotics as a contested domain of policy and practice in a comprehensive welfare state. We also highlight some benefits of this approach for analyzing media discourse and stakeholder perspectives on controversial technologies.

Gender stereotypes embedded in natural language are stronger in more economically developed and individualistic countries.

Gender stereotypes contribute to gender imbalances, and analyzing their variations across countries is important for understanding and mitigating gender inequalities. However, measuring stereotypes is difficult, particularly in a cross-cultural context. Word embeddings are a recent useful tool in natural language processing permitting to measure the collective gender stereotypes embedded in a society. In this work, we used word embedding models pre-trained on large text corpora from more than 70 different countries to examine how gender stereotypes vary across countries. We considered stereotypes associating men with career and women with family as well as those associating men with math or science and women with arts or liberal arts. Relying on two different sources (Wikipedia and Common Crawl), we found that these gender stereotypes are all significantly more pronounced in the text corpora of more economically developed and more individualistic countries. Our analysis suggests that more economically developed countries, while being more gender equal along several dimensions, also have stronger gender stereotypes. Public policy aiming at mitigating gender imbalances in these countries should take this feature into account. Besides, our analysis sheds light on the "gender equality paradox," i.e. on the fact that gender imbalances in a large number of domains are paradoxically stronger in more developed/gender equal/individualistic countries.

Identification of social scientifically relevant topics in an interview repository: a natural language processing experiment

PurposeThe present study is about generating metadata to enhance thematic transparency and facilitate research on interview collections at the Research Documentation Centre, Centre for Social Sciences (TK KDK) in Budapest. It explores the use of artificial intelligence (AI) in producing, managing and processing social science data and its potential to generate useful metadata to describe the contents of such archives on a large scale.Design/methodology/approachThe authors combined manual and automated/semi-automated methods of metadata development and curation. The authors developed a suitable domain-oriented taxonomy to classify a large text corpus of semi-structured interviews. To this end, the authors adapted the European Language Social Science Thesaurus (ELSST) to produce a concise, hierarchical structure of topics relevant in social sciences. The authors identified and tested the most promising natural language processing (NLP) tools supporting the Hungarian language. The results of manual and machine coding will be presented in a user interface.FindingsThe study describes how an international social scientific taxonomy can be adapted to a specific local setting and tailored to be used by automated NLP tools. The authors show the potential and limitations of existing and new NLP methods for thematic assignment. The current possibilities of multi-label classification in social scientific metadata assignment are discussed, i.e. the problem of automated selection of relevant labels from a large pool.Originality/valueInterview materials have not yet been used for building manually annotated training datasets for automated indexing of scientifically relevant topics in a data repository. Comparing various automated-indexing methods, this study shows a possible implementation of a researcher tool supporting custom visualizations and the faceted search of interview collections.

Welcome to the University of life, can I take your order? Investigating Life Experience Degree Offerings in Diploma mills

In this study, we posit a new category of products provided by diploma mills, which we term Life Experience Degree Offerings (LEDOs). LEDOs uniquely capitalise on the misuse of the principle of Accreditation of Prior Learning (APL) by granting higher education qualifications based on a resume or CV alone. Through a comparative analysis with contract cheating websites, we highlight the key features and persuasive strategies employed by 10 diploma mill websites which provide LEDOs to attract and convince potential clients. We then use corpus linguistics methods by analysing a large corpus of text quantitatively to explore the keywords these websites use to describe their products using a pre-packaged corpus analysis tool (Sketch Engine). Our findings indicate that on providers’ websites, the LEDOs are framed as a tool to achieve greater socioeconomic opportunities, and the quality and appearance of the physical diploma and accompanying documents play a key role in the LEDOs’ value proposition, while references to the acquisition of knowledge and process of learning are absent. Furthermore, LEDOs are typified by the promise of accreditation and verification services, which are the two most common keywords used in the description of LEDOs on diploma mills’ websites. Future research directions are suggested, including examining this phenomenon in non-Western cultural contexts, understanding the users and operators of LEDO websites, and assessing the prevalence of fraudulent credentials obtained from these sites.

Using Machine Learning of Online Expression to Explain Recovery Trajectories: Content Analytic Approach to Studying a Substance Use Disorder Forum.

Smartphone-based apps are increasingly used to prevent relapse among those with substance use disorders (SUDs). These systems collect a wealth of data from participants, including the content of messages exchanged in peer-to-peer support forums. How individuals self-disclose and exchange social support in these forums may provide insight into their recovery course, but a manual review of a large corpus of text by human coders is inefficient. The study sought to evaluate the feasibility of applying supervised machine learning (ML) to perform large-scale content analysis of an online peer-to-peer discussion forum. Machine-coded data were also used to understand how communication styles relate to writers' substance use and well-being outcomes. Data were collected from a smartphone app that connects patients with SUDs to online peer support via a discussion forum. Overall, 268 adult patients with SUD diagnoses were recruited from 3 federally qualified health centers in the United States beginning in 2014. Two waves of survey data were collected to measure demographic characteristics and study outcomes: at baseline (before accessing the app) and after 6 months of using the app. Messages were downloaded from the peer-to-peer forum and subjected to manual content analysis. These data were used to train supervised ML algorithms using features extracted from the Linguistic Inquiry and Word Count (LIWC) system to automatically identify the types of expression relevant to peer-to-peer support. Regression analyses examined how each expression type was associated with recovery outcomes. Our manual content analysis identified 7 expression types relevant to the recovery process (emotional support, informational support, negative affect, change talk, insightful disclosure, gratitude, and universality disclosure). Over 6 months of app use, 86.2% (231/268) of participants posted on the app's support forum. Of these participants, 93.5% (216/231) posted at least 1 message in the content categories of interest, generating 10,503 messages. Supervised ML algorithms were trained on the hand-coded data, achieving F1-scores ranging from 0.57 to 0.85. Regression analyses revealed that a greater proportion of the messages giving emotional support to peers was related to reduced substance use. For self-disclosure, a greater proportion of the messages expressing universality was related to improved quality of life, whereas a greater proportion of the negative affect expressions was negatively related to quality of life and mood. This study highlights a method of natural language processing with potential to provide real-time insights into peer-to-peer communication dynamics. First, we found that our ML approach allowed for large-scale content coding while retaining moderate-to-high levels of accuracy. Second, individuals' expression styles were associated with recovery outcomes. The expression types of emotional support, universality disclosure, and negative affect were significantly related to recovery outcomes, and attending to these dynamics may be important for appropriate intervention.

The (moral) language of hate.

Humans use language toward hateful ends, inciting violence and genocide, intimidating and denigrating others based on their identity. Despite efforts to better address the language of hate in the public sphere, the psychological processes involved in hateful language remain unclear. In this work, we hypothesize that morality and hate are concomitant in language. In a series of studies, we find evidence in support of this hypothesis using language from a diverse array of contexts, including the use of hateful language in propaganda to inspire genocide (Study 1), hateful slurs as they occur in large text corpora across a multitude of languages (Study 2), and hate speech on social-media platforms (Study 3). In post hoc analyses focusing on particular moral concerns, we found that the type of moral content invoked through hate speech varied by context, with Purity language prominent in hateful propaganda and online hate speech and Loyalty language invoked in hateful slurs across languages. Our findings provide a new psychological lens for understanding hateful language and points to further research into the intersection of morality and hate, with practical implications for mitigating hateful rhetoric online.

An overview of the consumer‐centric disruptive technology research: Insights from topic modelling and literature review

AbstractThis review aims to assess the most prevalent disruptive technologies of the past decade and evaluate them from a dual perspective. This study utilises the technology life cycle framework to classify disruptive technologies into three phases, (i) emerging, (ii) industry, and (iii) customer‐focused research outcomes. Subsequently, entities using six thematic categories via topic modelling are assessed to determine the research trajectory about customers, industries, and theory. The present study utilises the Latent Dirichlet Allocation framework for topic modelling, a quantitative tool that employs advanced statistical methodology to extract semantic knowledge from large text corpora. This approach offers a more comprehensive and insightful categorisation of information compared to traditional research review models such as archive review, meta‐review analysis, and systematic literature review. After the review, over 750 articles were analysed and classified according to their unique themes and technological emphasis. The study suggests practical applications that could enhance industrial progress through disruptive technology. Furthermore, by identifying gaps, the authors suggest multiple research prospects for forthcoming investigations on disruptive technologies and their impact on consumer behaviour. This research is novel in various aspects. This research is by far the only research record in the last decade that provides a systematic literature review of disruptive technologies with an extensive focus on customers. Considering the methodological perspective, this research uses topic modelling as a new literature review model, enabling analysing more research outputs and yielding more valuable findings than traditional methodologies.

Annotation uncertainty in the context of grammatical change

Abstract This paper elaborates on the notion of uncertainty in the context of annotation in large text corpora, specifically focusing on (but not limited to) historical languages. Such uncertainty might be due to inherent properties of the language, for example, linguistic ambiguity and overlapping categories of linguistic description, but could also be caused by a lack of annotation expertise. By examining annotation uncertainty in more detail, we identify the sources, deepen our understanding of the nature and different types of uncertainty encountered in daily annotation practice, and discuss practical implications of our theoretical findings. This paper can be seen as an attempt to reconcile the perspectives of the main scientific disciplines involved in corpus projects, linguistics and computer science, to develop a unified view and to highlight the potential synergies between these disciplines.

Keyword/ Keyphrase Extraction from Text of Indian Election Domain

In recent years, there has been a tremendous increase in the amount of data generated from various sources including social media, news articles, and blogs. With the rise of social media platforms, people are expressing their opinions more freely than ever before. This has led to an explosion of data in the Indian Election domain, where people express their views on various political parties and candidates. In order to extract meaningful information from this vast amount of data, it is important to identify and extract relevant keywords and phrases. Keyword and phrase extraction is the process of automatically identifying important words and phrases from a piece of text. This process is crucial for various natural language processing tasks such as text mining, sentiment analysis, topic modeling, and text classification. In this research paper, we focus on the task of keyword and phrase extraction from Indian Election domain text. We aim to extract relevant keywords and phrases that are most commonly used in the context of Indian elections. This research is important as it can help in understanding the key issues and concerns of Indian voters during the election season. We use various natural language processing techniques and machine learning algorithms to extract keywords and phrases from a large corpus of Indian Election domain text. Our approach involves pre-processing the text, including tokenization, stop-word removal, stemming, and POS tagging. We then use various statistical and machine learning models to identify the most relevant keywords and phrases.

Regina Coeli—Doctrine and Iconography of the Virgin Mary’s Heavenly Royalty

This article aims to highlight the privileged status granted by Christianity to the Virgin Mary when considering her the Queen of Heaven. From the very early centuries of our era, this sublime title was assigned to the Virgin Mary, for her condition of Mother of God, by an increasing number of Church Fathers and theologians. Later, it was expounded by an uncountable number of medieval liturgical hymns. Finally, from the 12th century onwards, these textual proclamations were shaped in sculptures and paintings according to various iconographic types. The author will proceed by first analyzing a large corpus of texts by many Church Fathers, theologians, and hymnographers. Then, he will consider twenty sculptures and paintings which reflect the heavenly royalty of Mary according to five iconographic types.

Exploiting Structure in Regular Expression Queries

Regular expression, or regex, is widely used to extract critical information from a large corpus of formatted text by finding patterns of interest. In tasks like log processing, the speed of regex matching is crucial. Data scientists and developers regularly use regex libraries that implement optimized regular expression matching using modern automata theory. However, computing state transitions in the underlying regex evaluation engine can be inefficient when a regex query contains a multitude of string literals. This inefficiency is further exasperated when analyzing large data volumes. This paper presents BLARE, Blazingly Fast Regular Expression, a regular expression matching framework that is inspired by the mechanisms that are used in database engines, which use a declarative framework to explore multiple equivalent execution plans, all of which produce the correct final result. Similarly, BLARE decomposes a regex into multiple regex and string components and then creates evaluation strategies in which the components can be evaluated in an order that is not strictly a left-to-right translation of the input regex query. Rather than using a cost-based optimization approach, BLARE uses an adaptive runtime strategy based on a multi-armed bandit approach to find an efficient execution plan. BLARE is also modular and can be built on top of any existing regex library. We implemented BLARE on four commonly used regex libraries, RE2, PCRE2, Boost Regex, and ICU Regex, and evaluated it using two production workloads and one open-source workload. BLARE was 1.6× to 3.7× faster than RE2 and 3.4× to 7.9× faster than Boost Regex. PCRE2 did not finish on one of the workloads, but on the remaining two workloads, BLARE improved the performance of PCRE2 by 3.1× to over 100×. For the open-source dataset, BLARE provided a speed up of 61.7× for ICU Regex. BLARE code is publicly available at https://github.com/mush-zhang/Blare.

Algorithms propagate gender bias in the marketplace—with consumers’ cooperation

AbstractRecent research shows that algorithms learn societal biases from large text corpora. We examine the marketplace‐relevant consequences of such bias for consumers. Based on billions of documents from online text corpora, we first demonstrate that from gender biases embedded in language, algorithms learn to associate women with more negative consumer psychographic attributes than men (e.g., associating women more closely with impulsive vs. planned investors). Second, in a series of field experiments, we show that such learning results in the delivery of gender‐biased digital advertisements and product recommendations. Specifically, across multiple platforms, products, and attributes, we find that digital advertisements containing negative psychographic attributes (e.g., impulsive) are more likely to be delivered to women compared to men, and that search engine product recommendations are similarly biased, which influences consumer's consideration sets and choice. Finally, we empirically examine consumer's role in co‐producing algorithmic gender bias in the marketplace and observe that consumers reinforce these biases by accepting gender stereotypes (i.e., clicking on biased ads). We conclude by discussing theoretical and practical implications.

San-Eng: Sanskrit to English Translator using Machine Learning

In the past decade, machine learning has made great strides in improving automatic translation. In general, machine learning algorithms have been able to achieve better translation quality by learning from large amounts of data. One approach to machine translation is to use a statistical machine translation (SMT) system. SMT systems learn to translate by statistical methods, using a large parallel corpus of text in multiple languages. The most successful SMT systems are based on the translation model known as the phrase-based translation model. In recent years, a new approach to machine translation has emerged, known as neural machine translation (NMT). NMT systems use artificial neural networks to learn to translate. NMT systems have shown to be very successful in translating between languages that are closely related, such as English and Hindi or Reverse is also true. In recent years, a new approach to machine translation has emerged, known as neural machine translation (NMT). NMT systems use artificial neural networks to learn to translate. NMT systems have shown to be very successful in translating between languages that are closely related, such as English and Hindi or Reverse is also true

The impact of big data on research methods in information science

Social media platforms, search engine transaction logs, Web portal transaction logs, large text corpora, and other data sources offer users a variety of invaluable data sources. Data generated based on social media platforms and other data sources have become important sources of big data for researchers in information science. Big data provide not only challenges but also opportunities for information science. Emerging big data trends inevitably have an impact on research methods in information science. The authors of this paper discuss the impact of big data on research methods in information science.This paper addresses these challenges and opportunities through the lens of research methods, ranging from data processing, to sampling, to information visualization, to temporal analysis, to sentiment analysis, to correlation, to cause-effect relationship, to data accessibility, to data privacy, and data ethics issues.The discussions on related aspects of research methods provoke a healthy reflection and debate for the information science research community, which can help researchers in the field produce sound research designs for big data-oriented research studies and assist information practitioners in solving problems they face in the big data age.Research methods are fundamental and essential for any research studies. Big data analysis has a natural relationship with research methods. The paper discusses an emerging and important research topic big data from the unique angle of research methods.

Automated Approach for Digitalizing Scope of Work Requirements to Support Contract Management

Scope of work requirements (SOWRs) specify the contractual obligations for the contractor to fulfill. Managing the information is crucial to the success of a construction project. This process is, however, often challenging because requirements are typically voluminous and written in an unstructured paper-based format. The current state-of-the-art models are mostly applicable to quantitative constraints for use in automated code compliance checking of design features. This study proposes a novel natural language processing (NLP) framework capable of digitalizing nonquantitative natural language SOWRs outlined in construction contracts. The model includes a comprehensive set of semantic and syntactic rules of linguistic features for extracting contractual work information including actors, actions, objects, constraints, tasks, and obligation. The framework was evaluated on a large textual corpus of provisions of design-build highway contracts. The model yields an impressive precision and recall of more than 93% and 87%, respectively. The proposed system is expected to help the project planner quickly develop an electronic database of contractual work requirements, enabling significantly improved efficiency in project planning and verification.

Predicting Generalized Anxiety Disorder From Impromptu Speech Transcripts Using Context-Aware Transformer-Based Neural Networks: Model Evaluation Study.

The ability to automatically detect anxiety disorders from speech could be useful as a screening tool for an anxiety disorder. Prior studies have shown that individual words in textual transcripts of speech have an association with anxiety severity. Transformer-based neural networks are models that have been recently shown to have powerful predictive capabilities based on the context of more than one input word. Transformers detect linguistic patterns and can be separately trained to make specific predictions based on these patterns. This study aimed to determine whether a transformer-based language model can be used to screen for generalized anxiety disorder from impromptu speech transcripts. A total of 2000 participants provided an impromptu speech sample in response to a modified version of the Trier Social Stress Test (TSST). They also completed the Generalized Anxiety Disorder 7-item (GAD-7) scale. A transformer-based neural network model (pretrained on large textual corpora) was fine-tuned on the speech transcripts and the GAD-7 to predict whether a participant was above or below a screening threshold of the GAD-7. We reported the area under the receiver operating characteristic curve (AUROC) on the test data and compared the results with a baseline logistic regression model using the Linguistic Inquiry and Word Count (LIWC) features as input. Using the integrated gradient method to determine specific words that strongly affect the predictions, we inferred specific linguistic patterns that influence the predictions. The baseline LIWC-based logistic regression model had an AUROC value of 0.58. The fine-tuned transformer model achieved an AUROC value of 0.64. Specific words that were often implicated in the predictions were also dependent on the context. For example, the first-person singular pronoun "I" influenced toward an anxious prediction 88% of the time and a nonanxious prediction 12% of the time, depending on the context. Silent pauses in speech, also often implicated in predictions, influenced toward an anxious prediction 20% of the time and a nonanxious prediction 80% of the time. There is evidence that a transformer-based neural network model has increased predictive power compared with the single word-based LIWC model. We also showed that the use of specific words in a specific context-a linguistic pattern-is part of the reason for the better prediction. This suggests that such transformer-based models could play a useful role in anxiety screening systems.

Review of Natural Language Processing in Pharmacology.

Natural language processing (NLP) is an area of artificial intelligence that applies information technologies to process the human language, understand it to a certain degree, and use it in various applications. This area has rapidly developed in the past few years and now employs modern variants of deep neural networks to extract relevant patterns from large text corpora. The main objective of this work is to survey the recent use of NLP in the field of pharmacology. As our work shows, NLP is a highly relevant information extraction and processing approach for pharmacology. It has been used extensively, from intelligent searches through thousands of medical documents to finding traces of adversarial drug interactions in social media. We split our coverage into five categories to survey modern NLP: methodology, commonly addressed tasks, relevant textual data, knowledge bases, and useful programming libraries. We split each of the five categories into appropriate subcategories, describe their main properties and ideas, and summarize them in a tabular form. The resulting survey presents a comprehensive overview of the area, useful to practitioners and interested observers. SIGNIFICANCE STATEMENT: The main objective of this work is to survey the recent use of NLP in the field of pharmacology in order to provide a comprehensive overview of the current state in the area after the rapid developments that occurred in the past few years. The resulting survey will be useful to practitioners and interested observers in the domain.

An Informed Neural Network for Discovering Historical Documentation Assisting the Repatriation of Indigenous Ancestral Human Remains

Among the pressing issues facing Australian and other First Nations peoples is the repatriation of the bodily remains of their ancestors, which are currently held in Western scientific institutions. The success of securing the return of these remains to their communities for reburial depends largely on locating information within scientific and other literature published between 1790 and 1970 documenting their theft, donation, sale, or exchange between institutions. This article reports on collaborative research by data scientists and social science researchers in the Research, Reconcile, Renew Network (RRR) to develop and apply text mining techniques to identify this vital information. We describe our work to date on developing a machine learning-based solution to automate the process of finding and semantically analysing relevant texts. Classification models, particularly deep learning-based models, are known to have low accuracy when trained with small amounts of labelled (i.e. relevant/non-relevant) documents. To improve the accuracy of our detection model, we explore the use of an Informed Neural Network (INN) model that describes documentary content using expert-informed contextual knowledge. Only a few labelled documents are used to provide specificity to the model, using conceptually related keywords identified by RRR experts in provenance research. The results confirm the value of using an INN network model for identifying relevant documents related to the investigation of the global commercial trade in Indigenous human remains. Empirical analysis suggests that this INN model can be generalized for use by other researchers in the social sciences and humanities who want to extract relevant information from large textual corpora.

Universal versus system-specific features of punctuation usage patterns in major Western languages

The celebrated proverb that “speech is silver, silence is golden” has a long multinational history and multiple specific meanings. In written texts punctuation can be considered one of its manifestations. Indeed, the virtue of effectively speaking and writing involves – often decisively – the capacity to apply the properly placed breaks. In the present study, based on a large corpus of world-famous and representative literary texts in seven major Western languages, it is shown that the distribution of intervals between consecutive punctuation marks can almost universally be characterized by only two parameters of the discrete Weibull distribution which can be given an intuitive interpretation in terms of the so-called hazard function. The values of these two parameters tend to be language-specific, however, and even appear to navigate translations. The properties of the computed hazard functions indicate that among the studied languages, English appears the least constrained by the necessity to place a consecutive punctuation mark to partition a sequence of words. Spanish reveals a similar tendency. The above characteristics manifest themselves when all consecutive punctuation marks are considered and not the sentence-ending ones only, and this fact is in accordance with the Detrended Fluctuation Analysis of intervals between the corresponding breaks.