Computational Linguistic Tools Research Articles

Computer Analysis of Scientific Texts to Identify Ideological Narratives

This article is devoted to the problem of identifying hidden ideological attitudes in scientific articles on economics, as well as defining methods for analyzing the ideological landscape of Russian economic science. The problem of this review article lies in the absence of a developed generally accepted approach to the textual analysis of scientific publications. Thus, there is an extremely wide list of methods that can be used in this case. Nevertheless, the development of the most effective methodology requires a comprehensive analysis of existing approaches individually and a comparison of each of them with the others. The authors associate the solution of the problem of choice in the presence of several approaches and a variety of methods with the definition of such a basis that would allow the formation of a methodology that includes ideological classification within the framework of economic scientific literature, the use of computational linguistics tools, in particular natural language processing, in addition to classical narrative analysis tools to identify "ideological narratives". In this regard, the methods of identifying implicit meanings in economic texts are considered in detail. The proposed tools can be used in further research on this or related topics.

Large Language Models as Computational Linguistics Tools: A Comparative Analysis of ChatGPT and Google Machine Translations

Large Language Models as Computational Linguistics Tools: A Comparative Analysis of ChatGPT and Google Machine Translations

Reform, Fighting Corruption, and Countering Terrorism by the Jordanian Media: A Corpus-Based Analysis of the Jordanian Local News Articles

This study combines the quantitative tools of computational linguistics and the qualitative methods of critical discourse analysis to examine the terms reform, corruption, and terrorismin Jordanian media before, during, and after the Arab Spring. The study builds three sub-corpora representative of the three periods under investigation: the pre-uprising corpus consists of 2059 news items from 2005 to 2009; the during-uprising corpus includes 2957 news items from 2011 to 2012; and the post-uprising corpus has 2436 news items from 2016 to 2017. In data analysis, the study uses the tools provided by AntConc software program to uncover the sociopolitical factors and challenges associated with reform, fightingcorruption, and counteringterrorism in Jordan. The findings of the study reveal that local news stories about terrorism, corruption, and reformdiffer significantly in terms of frequency and content in the pre-, post-, and during-uprising periods. The study indicates that sociopolitical conditions influence news coverage of the topics reform, corruption, and terrorismand thatnews articles are the medium where socialand political events are manifested. Keywords: Corpus linguistics, discourse analysis, Jordanian newspaper, media, Arab Spring

Lingualyzer: A computational linguistic tool for multilingual and multidimensional text analysis

Most natural language models and tools are restricted to one language, typically English. For researchers in the behavioral sciences investigating languages other than English, and for those researchers who would like to make cross-linguistic comparisons, hardly any computational linguistic tools exist, particularly none for those researchers who lack deep computational linguistic knowledge or programming skills. Yet, for interdisciplinary researchers in a variety of fields, ranging from psycholinguistics, social psychology, cognitive psychology, education, to literary studies, there certainly is a need for such a cross-linguistic tool. In the current paper, we present Lingualyzer (https://lingualyzer.com), an easily accessible tool that analyzes text at three different text levels (sentence, paragraph, document), which includes 351 multidimensional linguistic measures that are available in 41 different languages. This paper gives an overview of Lingualyzer, categorizes its hundreds of measures, demonstrates how it distinguishes itself from other text quantification tools, explains how it can be used, and provides validations. Lingualyzer is freely accessible for scientific purposes using an intuitive and easy-to-use interface.

Pedagogical discourse markers in online algebra learning: Unraveling instructor's communication using natural language processing

Despite the proliferation of video-based instruction and its benefits—such as promoting student autonomy and self-paced learning—the complexities of online teaching remain a challenge. To be effective, educators require extensive training in digital teaching methodologies. As such, there's a pressing need to examine and comprehend the intricacies of instructors' communication patterns within this context. This research addresses the pressing need to understand pedagogical discourse in online video lectures in Algebra classes by employing computational linguistic tools and natural language processing (NLP). Using transcripts from 125 Algebra 1 video lectures—comprising 4962 instances of pedagogical discourse—from five instructors at Math Nation, a virtual math learning environment, we analyzed the conveyance of linguistic, attitudinal, and emotional nuances. With the aid of 26 Coh-Metrix and SÉANCE features, we classified educators' language choices, achieving an accuracy of 86.7%. Furthermore, variations in language choices, as signified by discourse markers, were examined through a K-means clustering approach. The resulting 17 clusters were grouped into interpersonal, structural, and cognitive pedagogic functions. Through this exploration, we demonstrate the promising potential of NLP in efficiently deciphering pedagogical communication patterns in video lectures. These insights open a new avenue for research, aimed at assessing the efficacy of digital instruction by scrutinizing pedagogical discourse characteristics in computer-based learning environments.

"The algorithm will screw you": Blame, social actors and the 2020 A Level results algorithm on Twitter.

In August 2020, the UK government and regulation body Ofqual replaced school examinations with automatically computed A Level grades in England and Wales. This algorithm factored in school attainment in each subject over the previous three years. Government officials initially stated that the algorithm was used to combat grade inflation. After public outcry, teacher assessment grades used instead. Views concerning who was to blame for this scandal were expressed on the social media website Twitter. While previous work used NLP-based opinion mining computational linguistic tools to analyse this discourse, shortcomings included accuracy issues, difficulties in interpretation and limited conclusions on who authors blamed. Thus, we chose to complement this research by analysing 18,239 tweets relating to the A Level algorithm using Corpus Linguistics (CL) and Critical Discourse Analysis (CDA), underpinned by social actor representation. We examined how blame was attributed to different entities who were presented as social actors or having social agency. Through analysing transitivity in this discourse, we found the algorithm itself, the UK government and Ofqual were all implicated as potentially responsible as social actors through active agency, agency metaphor possession and instances of passive constructions. According to our results, students were found to have limited blame through the same analysis. We discuss how this builds upon existing research where the algorithm is implicated and how such a wide range of constructions obscure blame. Methodologically, we demonstrated that CL and CDA complement existing NLP-based computational linguistic tools in researching the 2020 A Level algorithm; however, there is further scope for how these approaches can be used in an iterative manner.

In the melting pot of web‐crawled texts: The challenges of extracting English words from Croatian corpora

AbstractThe focus of this paper are English words and phrases used in Croatian which, unlike loanwords, have not undergone major adaptations at the orthographic, phonetic, or other levels apart from being influenced by the inflectional system of the recipient language. A list of English words in Croatian corpora was compiled using automatic algorithm extraction, corpus query language in Sketch Engine, and manual word list evaluation with the end goal of publishing the first comprehensive online database of English words in Croatian. The ENGRI corpus of Croatian was created by web crawling procedure and used together with the existing Croatian hrWaC 2.2 RFTagger corpus to produce a list of English words and phrases. In this paper, word list compilation issues are discussed in relation to both general issues encountered in the study of interlingual lexical types (such as false cognates, antonomasia, and polysemy) as well as Croatian‐specific language properties such as its inflectional system and diacritical marks. In conclusion, we propose that manual evaluation is an indispensable method and a necessary complement to computational linguistic tools in the creation of word lists and databases of foreign words in other languages.

Novice Russian Research Writing: Prepositions and Prepositional Phrases

Background: The research writing of novice Russian authors tend to be markedly different from that of expert academic writers from other countries. More specifically, Russian student writing has been characterized as wordy, difficult to comprehend, syntactically complex, and excessive in terms of nominalisation. One of the main manifestations of these characteristics is the deployment of a large number of prepositions and prepositional phrases.
 Purpose: The purpose of this paper is to investigate the causes of this excessive use of prepositions in Russian student writing and to provide suggestions for improvement.
 Methods: The quantitative analysis evaluates two self-compiled corpora using the computational linguistics tool Gramulator. The first corpus consists of published research papers written by international scholars of radio engineering. The second corpus comprises first drafts of research papers written by Russian graduate and postgraduate students majoring in radio engineering. The final qualitative analysis focuses largely on the student corpus.
 Results: The seven most common writing features identified were as follows: excessive of-phrases, nouns/verbal nouns instead of gerunds, nouns instead of infinitives of purpose, nominalized structures instead of relative clauses, ‘strong noun + weak verb’ structures instead of ‘strong’ verbs, grammatical errors, and repetitions. Each of these features is discussed and followed by suggestions that may help both reduce the excessive number of prepositions and prepositional phrases and improve other important features of the text.
 Implications: The results of this study are of interest to academic writing instructors as well as the developers of teaching materials and automated evaluation tools.

Corpus-Based Analysis of French and English Neologisms in Media Discourse

Language development is an ongoing process, and changes occur at all levels of the lan-guage. This article discusses the changes in the lexical composition of French and English, namely neolo-gisms. Media discourse was chosen as the material for the study because it is characterized by a high dyna-mism of changes and reflects, interprets many new phenomena of objective reality, with new language struc-tures used to describe them. The study was based on linguistic data from Néoveille and WebCorp platforms. In addition to quantitative and distributive methods, a qualitative and component analysis of neologisms was applied to their word-formation models and connotations. In 2022 the most frequent neologisms in French were personnes-années, demies-journées, grande-reporters, and tout-mots, in English –liveaboard, micro-mobility, effective altruism, and tomato flu. Additionally, some neologisms were analyzed in context. In 2022 metaphorical neologisms were found to be more common in English than in French. Neologismsap-pear more often in contexts that are related to technology and politics. Based on the study, lists of the most common neologisms were compiled, which can be used in the teaching of foreign languages at high levels. Undoubtedly, it is necessary to further study neologisms that appear in different languages and the ways of detecting them in various texts because, despite the high level of development of the tools of computational and corpus linguistics, their methods and tools still have limitations and shortcomings that require an expert assessment of the results obtained.

Widespread cortical thinning, excessive glutamate and impaired linguistic functioning in schizophrenia: A cluster analytic approach.

IntroductionSymptoms of schizophrenia are closely related to aberrant language comprehension and production. Macroscopic brain changes seen in some patients with schizophrenia are suspected to relate to impaired language production, but this is yet to be reliably characterized. Since heterogeneity in language dysfunctions, as well as brain structure, is suspected in schizophrenia, we aimed to first seek patient subgroups with different neurobiological signatures and then quantify linguistic indices that capture the symptoms of “negative formal thought disorder” (i.e., fluency, cohesion, and complexity of language production).MethodsAtlas-based cortical thickness values (obtained with a 7T MRI scanner) of 66 patients with first-episode psychosis and 36 healthy controls were analyzed with hierarchical clustering algorithms to produce neuroanatomical subtypes. We then examined the generated subtypes and investigated the quantitative differences in MRS-based glutamate levels [in the dorsal anterior cingulate cortex (dACC)] as well as in three aspects of language production features: fluency, syntactic complexity, and lexical cohesion.ResultsTwo neuroanatomical subtypes among patients were observed, one with near-normal cortical thickness patterns while the other with widespread cortical thinning. Compared to the subgroup of patients with relatively normal cortical thickness patterns, the subgroup with widespread cortical thinning was older, with higher glutamate concentration in dACC and produced speech with reduced mean length of T-units (complexity) and lower repeats of content words (lexical cohesion), despite being equally fluent (number of words).ConclusionWe characterized a patient subgroup with thinner cortex in first-episode psychosis. This subgroup, identifiable through macroscopic changes, is also distinguishable in terms of neurochemistry (frontal glutamate) and language behavior (complexity and cohesion of speech). This study supports the hypothesis that glutamate-mediated cortical thinning may contribute to a phenotype that is detectable using the tools of computational linguistics in schizophrenia.

Firms as Influencers: Shaping Industries through Search and Collaboration

Firms often engage in shaping the various dimensions of an industry's competitive landscape. While the actions by which firms are able to shape an industry has received increased attention, research has yet to examine the knowledge antecedents of firms’ shaping behavior; that is the knowledge that allows firms to create a persistent change to the competitive landscape. This paper takes a first step at quantitatively examining the information processing antecedents of industry shaping by employing a novel methodology using computational linguistic tools to provide evidence on how firms can engage in industry shaping. Specifically, I examine how a firm’s shaping of the positioning space depends on its depth and breadth knowledge search and the extent to which they collaborate with other firms. I study this question using machine learning techniques in combination with 2.4 million USPTO patents and 585,133 firm descriptions. The results show that search breadth and depth are positively associated with industry shaping while the effect of collaboration is contingent on a firm's level of breadth and depth. Overall, the findings suggest that depth enables a structural understanding of the environment allowing other firms to learn from the focal firm in collaborative relationships, enhancing their credibility. Meanwhile, breadth lowers the level of expertise, decreasing the amount of useful knowledge that other firms learn from collaboration.

Addressing discourse differences in the writing of Russian engineering students and international researchers

Virtually all researchers understand the requirement of presenting their studies in peer-reviewed English-medium journals. Russian scientific writers understand this necessity too; however, evidence suggests that these particular researchers are under-performing relative to similar non-native English speakers. The considerable challenge Russians face centers on articulating their ideas in English in the way that meets the norms and expectations of their international discourse community. That is, their writing is characterized as wordy, cumbersome, too academese, and syntactically complex. These issues need to be addressed in the early stages of developing writing skills. In this study, we address discourse differences between the scientific writing of Russian engineering students and that of international experts. Using the computational linguistics tools Coh-Metrix and Gramulator, we compare a corpus of students’ manuscripts with a similar corpus of experts’ published papers. We focused on six conceptual categories: readability, writing quality, cohesion, syntax, word choice, and genre purity. The overall results suggest that student writing differs significantly for multiple characteristics of text and discourse. A discriminant analysis provided a model that successfully predicts group membership and helps identify the most important issues in Russian student writing. Measures such as noun phrase density, genre purity, word age of acquisition, and variance in sentence length were found to be significant positive predictors of Russian student writing, whereas lexical diversity, adversative/contrastive connectives, adverbial phrase density, and word concreteness were significant positive predictors of expert writing. Our analysis allowed us to provide guidance for instructors, materials designers, as well as for technological assessment tools.

A Language-Based Method for Assessing Symbolic Boundary Maintenance between Social Groups

When the social boundaries between groups are breached, the tendency for people to erect and maintain symbolic boundaries intensifies. Drawing on extant perspectives on boundary maintenance, we distinguish between two strategies that people pursue in maintaining symbolic boundaries: boundary retention—entrenching themselves in pre-existing symbolic distinctions—and boundary reformation—innovating new forms of symbolic distinction. Traditional approaches to measuring symbolic boundaries—interviews, participant-observation, and self-reports are ill-suited to detecting fine-grained variation in boundary maintenance. To overcome this limitation, we use the tools of computational linguistics and machine learning to develop a novel approach to measuring symbolic boundaries based on interactional language use between group members before and after they encounter one another. We construct measures of boundary retention and reformation using random forest classifiers that quantify group differences based on pre- and post-contact linguistic styles. We demonstrate this method's utility by applying it to a corpus of email communications from a mid-sized financial services firm that acquired and integrated two smaller firms. We find that: (a) the persistence of symbolic boundaries can be detected for up to 18 months after a merger; (b) acquired employees exhibit more boundary reformation and less boundary retention than their counterparts from the acquiring firm; and (c) individuals engage in more boundary retention, but not reformation, when their local work environment is more densely populated by ingroup members. We discuss implications of these findings for the study of culture in a wide range of intergroup contexts and for computational approaches to measuring culture.

Aligning Differences: Discursive Diversity and Team Performance

How does cognitive diversity in a group affect its performance? Prior research suggests that group cognitive diversity poses a performance tradeoff: Diverse groups excel at creativity and innovation, but struggle to take coordinated action. Building on the insight that group cognition is not static, but is instead dynamically and interactively produced, we introduce the construct of discursive diversity, a manifestation of group cognitive diversity, which reflects the degree to which the meanings conveyed by group members in a given set of interactions diverge from one another. We propose that high-performing teams are ones that have a collective capacity to modulate shared cognition to match changing task requirements: They exhibit higher discursive diversity when engaged in ideational tasks and lower discursive diversity when performing coordination tasks. We further argue that teams exhibiting congruent modulation—that is, those with low group-level variance in members’ within-person semantic shifts to changing task requirements—are more likely to experience success than teams characterized by incongruent modulation. Using the tools of computational linguistics to derive a measure of discursive diversity and drawing on a novel longitudinal data set of intragroup electronic communications and performance outcomes for 117 remote software development teams on an online platform ( www.gigster.com ), we find support for our theory. Our findings suggest that the performance tradeoff of group cognitive diversity is not inescapable: Groups can navigate it by aligning their levels of discursive diversity to match their task requirements and by having members stay aligned with one another as they make these adjustments. This paper was accepted by Isabel Fernandez-Mateo, organizations. Funding: Financial support fromthe NSF-CAREER [Grant 1847091] is gratefully acknowledged. Supplemental Material: Data are available at https://doi.org/10.1287/mnsc.2021.4274 .

Mlphon: A Multifunctional Grapheme-Phoneme Conversion Tool Using Finite State Transducers

In this article we present the design and the development of a knowledge based computational linguistic tool, Mlphon [em.el.foːɳ] for Malayalam language. Mlphon computationally models linguistic rules using finite state transducers and performs multiple functions including grapheme to phoneme (g2p) and phoneme to grapheme (p2g) conversions, syllabification, phonetic feature analysis and script grammar check. This open source software tool, released under MIT license, is developed as a one-stop solution to handle different speech related text processing tasks for automatic speech recognition, text to speech synthesis and non-speech natural language processing tasks including syllable subword based language modeling, phoneme diversity analysis and text sanity check. The tool is evaluated on a manually crafted gold standard lexicon. Mlphon performs orthographic syllabification with 99% accuracy with a syllable error rate of 0.62% on the gold standard lexicon. For grapheme to phoneme conversion task, overall phoneme recognition accuracy of 99% with a phoneme error rate of 0.55% is obtained on gold standard lexicon. Additionally an extrinsic evaluation of Mlphon is performed by employing the pronunciation lexicon created using Mlphon, in Malayalam automatic speech recognition (ASR) task. Performance analysis in terms of the computation time of lexicon creation process and the word error rate (WER) on ASR task are presented along with a comparison over other automated tools for lexicon creation. Pronunciation lexicons with more than 100k commonly used Malayalam words in phonemised and syllabified forms is created and they are published as open language resources along with this work. We also demonstrate the usage of Mlphon on different natural language processing applications - syllable subword ASR, assisted pronunciation learning, phoneme diversity analysis and text sanity check. Being a knowledge based solution with open source code, Mlphon can be adapted to other languages of similar script nature.

Intellectual capital reporting and mandatory adoption of post-IFRS – An empirical analysis using computational linguistic tools

PurposeThe study aims to report empirical evidence on the impact of mandatory adoption of International Financial Reporting Standards (IFRS) in India on the voluntary intellectual capital reporting (ICR) and its value relevance. The study also tests the effect of term-weighting schemes used for information retrieval studies in the domain area of ICR.Design/methodology/approachThe study uses computational linguistics tools to measure ICR by Indian firms in the period 2014–2019. The study developed term frequencies for 23 ICR attributes using bag-of-words methodology from the annual reports. The word counts were used to construct two distinct measures of ICR, quantity and quality, deploying different term-weighting schemes, equal weighting and the term frequency-inverted document frequency (TF-IDF) weighting, respectively. A combination of parametric and non-parametric tests has been employed to examine the different hypothesis.FindingsThe quantity of ICR was found to have increased post-IFRS adoption. However, the quality of ICR had fallen significantly, which resulted in the loss of value relevance of ICR. Firms making higher disclosures but of inferior quality experienced suboptimal market returns. Variation in inter-firm ICR has reduced. Size effect and sector effect continue but have attenuated. The study acknowledges the enormous impact of term-weighting schemes, used for information retrieval studies, in the domain area of ICR.Practical implicationsThe study strongly adds to the momentum in favour of a formal ICR standard to improve its quality, restore its value relevance and facilitate more effective decision-making where the valuation of a firm is a critical input. The study presages the firms not to make poor-quality disclosures to avoid suboptimal stock performance.Originality/valueThe study sheds light on the impact of the adoption of post-IFRS on ICR in India. The study establishes the effect of term-weighting schemes, used for linguistic studies, in the domain area of ICR and adds to the literature by explaining one of the critical reasons for the dichotomy in ICR trends.

Suitability of Text-Based Communications for the Delivery of Psychological Therapeutic Services to Rural and Remote Communities: Scoping Review.

BackgroundPeople living in rural and remote areas have poorer access to mental health services than those living in cities. They are also less likely to seek help because of self-stigma and entrenched stoic beliefs about help seeking as a sign of weakness. E-mental health services can span great distances to reach those in need and offer a degree of privacy and anonymity exceeding that of traditional face-to-face counseling and open up possibilities for identifying at-risk individuals for targeted intervention.ObjectiveThis scoping review maps the research that has explored text-based e-mental health counseling services and studies that have used language use patterns to predict mental health status. In doing so, one of the aims was to determine whether text-based counseling services have the potential to circumvent the barriers faced by clients in rural and remote communities using technology and whether text-based communications, in particular, can be used to identify individuals at risk of psychological distress or self-harm.MethodsWe conducted a comprehensive electronic literature search of PsycINFO, PubMed, ERIC, and Web of Science databases for articles published in English through November 2020.ResultsOf the 9134 articles screened, 70 met the eligibility criteria and were included in the review. There is preliminary evidence to suggest that text-based, real-time communication with a qualified therapist is an effective form of e-mental health service delivery, particularly for individuals concerned with stigma and confidentiality. There is also converging evidence that text-based communications that have been analyzed using computational linguistic techniques can be used to accurately predict progress during treatment and identify individuals at risk of serious mental health conditions and suicide.ConclusionsThis review reveals a clear need for intensified research into the extent to which text-based counseling (and predictive models using modern computational linguistics tools) may help deliver mental health treatments to underserved groups such as regional communities, identify at-risk individuals for targeted intervention, and predict progress during treatment. Such approaches have implications for policy development to improve intervention accessibility in at-risk and underserved populations.

Decision-making and computational linguistic tools application for overall estimation of the level of social tension

The study of social tension, its causes and laws of variation, as well as establishing its level is crucial during crises, epidemics and wars. The purpose of this research is to develop the system for determining the level of tension in society on the basis of data from social networks using decision-making and computational linguistic tools. The impact of various factors on the growth of social tensions is determined with the help of knowledge bases built by means of decision support systems. To analyse the emotionality of comment on news publications the TF-IDF and Word2vec methods are used. Within the framework of the method of goal dynamic estimating of alternatives, the level of social tension is determined as the degree of achievement of the main goal, the ratings of news publications and the ratings of news events that contributed to the increase in social tension are calculated. The study presents the efficient method of determining the level of social tension and defining the main factors that have the greatest impact on the growth of social tension. Further studies are needed to improve the estimates of factors and the overall estimate of social tension.

The lexical analysis of two works by Ernest Hemingway and F. Scott Fitzgerald

The scientific field of computational linguistics can significantly contribute to the analysis of literary texts from a variety of perspectives, including educational ones. The purpose of the present study is to investigate and analyze literary texts with the help of computational linguistics devices, with special focus on the difficulty level of vocabulary items, the general vocabulary profile analysis regarding the frequency of occurrence, and the sentence length of selected texts, on which research questions of the present study are based. Ernest Hemingway’s work Big Two-Hearted River – Part I., and F. Scott Fitzgerald’s The Great Gatsby, were analyzed. It is hypothesized, based on previous research, that the words used in Hemingway’s prose will fall into a lower reading difficulty range than that of Fitzgerald’s, and that Hemingway’s style will generally be simpler than that of Fitzgerald’s in terms of syntactic structures and sentence length. Their writings were analyzed with the help of computational linguistics tools. Comparing the text profiles of Hemingway and Fitzgerald for the level of difficulty, it can be concluded that the vocabulary level of their writings is not significantly different. Yet, Fitzgerald’s prose contains significantly longer and more elaborate sentences. Language technology appliances may contribute to the critical, detailed and effective analysis of literary works, contributing to other benefits, including language teaching.

Text mining tool for translating terms of contract into technical specifications: Development and application in the railway sector

Tenders or technical terms contain a large quantity of both technical, legal, managerial information mixed in a nested and complex net of relationships. Extracting technical and design information from a document whose aim is both legal and technical, and that is written using several specific jargons, is not a trivial task: the purpose of the research is to try to detect, extract, split and assign information from the text of a tender in an automatic way. It means being able to understand technical and legal terms and organize them in multiple ways: according to product structure, internal organisational structure, etc. The focus is in providing a handy tool that could speed up and facilitate human analysis and allow tackling also the process of transforming customer’s requirements into design specifications. The approach chosen to overcome the various issues is to support state-of-the-art Computational Linguistic tools with a wide Knowledge Base. The latter has been constructed both manually and automatically and comprises not only keywords but also concepts, relationships and regular expressions. The implementation of the methodology has been carried out during a project for AnsaldoBreda S.p.A. (now Hitachi Rail Europe). A case study about the tender for a high-speed train has been included to show the functioning and output of the entire software system.