Text Analysis Techniques Research Articles

Supervised probabilistic latent semantic analysis with applications to controversy analysis of legislative bills

Probabilistic Latent Semantic Analysis (PLSA) is a fundamental text analysis technique that models each word in a document as a sample from a mixture of topics. PLSA is the precursor of probabilistic topic models including Latent Dirichlet Allocation (LDA). PLSA, LDA and their numerous extensions have been successfully applied to many text mining and retrieval tasks. One important extension of LDA is supervised LDA (sLDA), which distinguishes itself from most topic models in that it is supervised. However, to the best of our knowledge, no prior work extends PLSA in a similar manner sLDA extends LDA by jointly modeling the contents and the responses of documents. In this paper, we propose supervised PLSA (sPLSA) which can efficiently infer latent topics and their factorized response values from the contents and the responses of documents. The major challenge lies in estimating a document’s topic distribution which is a constrained probability that is dictated by both the content and the response of the document. To tackle this challenge, we introduce an auxiliary variable to transform the constrained optimization problem to an unconstrained optimization problem. This allows us to derive an efficient Expectation and Maximization (EM) algorithm for parameter estimation. Compared to sLDA, sPLSA converges much faster and requires less hyperparameter tuning, while performing similarly on topic modeling and better in response factorization. This makes sPLSA an appealing choice for latent response analysis such as ranking latent topics by their factorized response values. We apply the proposed sPLSA model to analyze the controversy of bills from the United States Congress. We demonstrate the effectiveness of our model by identifying contentious legislative issues.

Living in a multicultural context: Health and integration from the perspective of undocumented Mediterranean migrants, residents, and stakeholders in Italy. A qualitative-multimethod study.

To catch a representative view of a multicultural population's needs. Qualitative study. Semi-structured interviews were conducted from July 2022 to January 2023 with the project's stakeholders, migrants, and residents. Data analysis was performed using a multimethod textual analysis technique. Territorial barriers, lack of social network, and specific professionals' training emerged as healthcare delivery obstacles. For migrants, language improvement emerged as a health priority. A deep relationship with migrants emerged as a deficiency for residents. A welcoming project equipped with solid leadership and the right resources can be fundamental in mediating health promotion and integration. In this process, the involvement of the resident population is essential. Analysing the migrants' needs and the strengths and limitations of a reception system could help identify the challenges for professionals in delivering culturally competent care. In this context, the nurse's role becomes relevant, being responsible for taking charge and caring for the population and the link between professionals and the population. WHAT PROBLEM DID THE STUDY ADDRESS?: The study addressed the problem of improving the overall health of migrants, refugees, and asylum seekers, mainly focusing on reception and integration into a new society process. WHAT WERE THE MAIN FINDINGS?: Worse health was identified with adaptation, integration, and family problems. Territorial barriers emerged, hindering good health. WHERE AND ON WHOM WILL THE RESEARCH HAVE AN IMPACT?: These research findings can be valuable for health professionals who want to improve the reception process and enhance a care model integrated with residents. No patient or public contribution. To describe the research report, we referred to the COREQ checklist (Tong etal., 2007).

Measuring the growth and impact of library and information science research in India: a multidimensional scientometric assessment

Purpose This study aims to provide insight into Library and Information Science (LIS) research in India using scientometric approaches. Web of Science (WoS) and SCOPUS databases were used for data retrieval. The study examines productivity in terms of source types, gender distribution, document formats, authorship and other factors. In addition, this study sought to identify trends or patterns in the research preferences of LIS scientists through text analysis. Design/methodology/approach Data were downloaded from the WoS and Scopus databases over 22 years and analysed using VOSviewer, Orange, Biblioshiny and CRExplorer softwares. Findings The findings reveal that 5,692 out of the 9,384 documents in both databases underwent the final examination. In total, 466 different sources produced all of those papers. Author analysis revealed that 6,603 different authors authored 5,692 documents. There were 4,209 male and 1,063 female authors. Furthermore, India shares maximum collaborations with the USA and England. The spectrogram features nine significant peaks corresponding to Lotka’s, Bradford’s and similar laws. Text analysis revealed that Indian LIS researchers have consistently investigated open access and digital or open libraries. Research limitations/implications The findings of this study will provide readers with a better understanding of India’s contribution to LIS. In addition, the study will help academics identify research gaps and undiscovered areas in the Indian context that require further investigation. Originality/value Not many studies highlight Indian research trends and international collaboration in LIS. This study highlights research trends, collaboration and gender productivity in LIS. The most cited references and trending topics were also identified using reference publication year spectroscopy and text analysis techniques.

Exploring social trends in cohousing and coliving discussions on X(Twitter) using NLP and Text Analysis Techniques

La investigación analiza las tendencias y variaciones en las discusiones relacionadas con cohousing y coliving en la red social X (anteriormente conocida como Twitter) entre 2019 y 2022. Utilizando técnicas avanzadas de análisis de redes de texto, la investigación utilizó Python y Snscrape para el preprocesamiento de texto, seguido por la construcción de gráficos de red y la detección de comunidades. El estudio emplea modelos de Asignación Latente de Dirichlet (LDA) para identificar los temas de discusión en los tweets y calcula el tf-idf de bigramas dentro de los principales clusters temáticos. Este estudio evalúa la importancia relativa de estos bigramas en función de su frecuencia en los documentos analizados. Los resultados revelan un patrón fractal de propagación de influencia dentro de la red social X. Temas clave como espacios de trabajo compartido, pisos en alquiler y planificación urbana destacan de manera prominente en las discusiones sobre la vivienda colaborativa, demostrando el impacto multifacético de los modelos de convivencia en diversas poblaciones. Esta investigación proporciona información esencial sobre el intrincado panorama de las conversaciones sobre la vivienda colaborativa, resaltando el papel fundamental de los modelos de convivencia en la atención de desafíos contemporáneos.

METHODS OF SUPPORTING ACCEPTANCE OF ECONOMICLY BASED DECISIONS IN THE MANAGEMENT OF THE QUALITY SYSTEM OF HIGHER EDUCATION

The aim is to influence the level of preparedness in both IT specialties and economic fields closely tied to IT technologies within higher or secondary educational institutions to meet the needs and requirements of diverse companies. Collaborative efforts between businesses and education can effectively tackle the challenge of cultivating high-quality potential personnel and enhancing students' readiness by establishing a standardized set of business demands.When outlining the requirements for IT specialties, crucial vectors encompass not only work experience and skills but also business qualities and the ability to collaborate in a work collective, aligning with company values. To precisely define the competencies of future employees, we advocate for the Delphi method as one of the best approaches. This method involves expert judgment and comprises multiple rounds of anonymous surveys within an expert group to reach a consensus.By fostering interactions between companies and educational institutions in the study, surveys can collect statistics and utilize educational plans that align with organizational interests. Proposed competencies by experts may convey similar semantic meanings but be articulated with different terminology. Additionally, merging similar competencies into more comprehensive ones is feasible. In the era of rapid IT development, addressing this issue can be accomplished through existing text analysis and comparison techniques in databases. Consequently, the unification of competencies can be established during the learning process.

Detecting cyberbullying in social media using text analysis and ensemble techniques

In the dynamic landscape of our hyper-connected digital world, social media platforms play a dual role as facilitators of global interaction and breeding grounds for harmful behaviors. Cyberbullying, an insidious online menace, inflicts emotional distress and psychological trauma on numerous individuals, underscoring the urgent need for advanced tools to detect and prevent such malevolent actions. This innovative project harnesses the power of artificial intelligence and text analysis to illuminate the dark corners of social media where cyberbullying thrives, offering hope to countless victims. At its core, this endeavor utilizes cutting-edge ensemble techniques, a fusion of diverse machine learning algorithms, to analyze textual content across social media platforms. This approach ensures unparalleled accuracy in identifying and flagging cyberbullying instances, enhancing the efficiency of the detection process while minimizing false positives. The project adopts a multifaceted approach to text analysis, examining explicit language, sentiments, context, and behavioral patterns in online interactions. By delving into the intricacies of human communication, the system distinguishes between genuine expressions and malicious intent, providing a nuanced and accurate assessment.

Validating Disaster Medicine Experts’ Feedback Outcome Using Artificial Intelligence Computing Techniques

Abstract Objectives This study aimed to use artificial intelligence (AI) computing techniques to determine if they can validate the findings of a previously published thematic analysis article focusing on disaster medicine experts’ open-ended feedback about Middle East and North African countries (MENA) for chemical, biological, radiological, and nuclear (CBRN) threats. Methods Automated text analytics techniques were employed to explore and visualize the semantic essence of the experts’ feedback through word vector transformation and Principal Component Analysis (PCA) for dimensionality reduction. The t-distributed Stochastic Neighbor Embedding (t-SNE) is another more advanced dimensionality reduction technique that enhanced the capturing of the determined components. Results Two prominent clusters emerged from the full textual data set representing word similarities groups in the original data set, denoting a thematic group of ideas that experts have emphasized in their responses. Upon deep reading the text feedback, the themes linked preparedness with different training types, such as tabletop exercises and policies/legislation. The findings are in line with currently adopted practices. Conclusions While AI methods demonstrated their valuable application in disaster medicine and helped validate the experts’ recommendations objectively, they should be approached cautiously, as they can be complex and challenging to comprehend fully.

The Relevance Of The Book On Marriage Schedules By The Mufti Of The Kingdom Of Kubu, West Kalimantan To Fikih Subjects At Madrasah Aliyah

The Book of Marriage Schedules is the great work of Mufti Ismail Mundu, a book full of knowledge, a book that talks about marriage. This article aims to explore and examine the contents of the book Jadual al-Nikāh and its relevance to the jurisprudence curriculum at Madrasah Aliyah. Writing with a qualitative approach, and a philological (History) approach, namely to obtain content information, the form of text presented by the author of the book, and bibliographic methods, with textual data analysis techniques for the Marriage Schedule book by condensing the data, presenting the data and drawing conclusions. This research resulted in an overview of the contents of the Marriage Schedule book on the one hand and the PAI curriculum, especially Mapel Fiqh on the other hand. This research concludes that the relationship between the discussion of marriage jurisprudence in the Marriage Schedule and the Jurisprudence Curriculum has quite high relevance. Thus, it is hoped that this book can be used as a reference as an additional learning resource in the fiqh learning process. Another function is that this book can be used as a learning medium in SKI learning using an Islamic history approach, as a learning medium in learning local content using a historical approach to the development of Islam in West Kalimantan based on local wisdom.

Exploring the surge of negativity during the COVID-19 pandemic: computational text and sentiment analysis across eight newsrooms’ tweets

ABSTRACT The rise of Twitter as a news platform has radically changed the way we access, consume, and share news. Twitter becomes an important hub to quickly and easily access accurate information in times of crisis such as COVID-19 and is frequently used in journalism practices. This study examines how the COVID-19 pandemic is covered by news agencies in the Twitter ecosystem, the weight of the news about the pandemic in tweets and the sentiment analysis of the news. Within the scope of the study, the tweets related to COVID-19 shared between 2020 and 2021 by eight news agencies (BBC World, Reuters, CNN, Associated Press, TRT World, AL Jazeera English, DW English, Euronews) that broadcast on a global scale and have a high number of followers on Twitter are analyzed by using text mining methods. Firstly, the frequently used words in tweets were obtained by using the text analysis technique n-gram. Secondly, the sentiment values of all the tweets and the words are computed and later classified into certain categories. Lexicon based sentiment dictionaries such as VADER and NRC utilized in the sentiment analysis process. Findings reveal that messages containing fear, anxiety, sadness, and negative polarity are prevalent in the news during the pandemic.

Anomaly detection in trade declarations using deep learning techniques: A risk-assessment approach to identify misclassification and incorrect valuation

In Hong Kong, merchandise trade statistics are compiled based on the commodity information given on the trade declarations submitted by traders. Due to the complexity of the standardised commodity classification system (i.e. Hong Kong Harmonized System, or HKHS in short), there are often reporting errors, especially in the commodity codes and quantities. With around 20 million declarations received annually, the availability of this big data source motivates us to adopt deep learning techniques to detect the reporting errors. This paper proposes a mechanism consisting of three deep learning models for checking the commodity code, quantity and value, which offers an end-to-end solution to data quality assurance for declarations. The results show that the proposed mechanism could enhance the accuracy of error detection, which is conducive to improving the quality of trade statistics. With the use of text analytics techniques, the mechanism could fully utilise free-text commodity descriptions declared by traders to check the accuracy of the declared information comprehensively. It also overcomes some limitations of the traditional rule-based models. The whole study demonstrates the potential of using deep learning approach in quality assurance of existing statistical systems for official statistics.

Organizational Traits and Social Badging: Text Mining Application

ABSTRACT This paper, employing a deductive research methodology approach, explores the concept of social badging for organizations from employees’ rating perspective in social media job portals (SMJP). This research attempts to identify the impact of monetary and non-monetary benefits as the antecedents and to explore the moderating effect of management support, organizational culture, and employee-perceived firm image on social badging in SMJPs. Relying on the signaling and person-environment fit theories, this study analyses more than 100,000 employee reviews from social media using text analytics techniques. The findings highlight the essential organizational traits to enhance favorable social badging for organizations. This study also has practical implications by underlying the critical drivers for reputational branding in social media.

Transmission mechanism of public concern in waste-sorting policy: Evidence from text mining

An escalating amount of urban waste poses a significant threat to the sustainable growth of cities. Therefore, a waste-sorting policy is crucial for the advancement of humankind. The implementation of a waste-sorting policy relies heavily on public participation to ensure effective governance. This study examines public participation in urban waste-sorting by mining more than 580,000 microblog texts related to waste-sorting from 2012 to 2022, using data from Sina Weibo, China's leading social media platform. My findings indicate that (1) residents’ attention to and support for mandatory waste-sorting policies varies by region; (2) widespread public interest stimulates publicity, education, and commercial entertainment related to waste-sorting policies and investment in thematically related sectors; (3) residents in regions with high education and income levels are more likely to generate waste and pay greater attention to waste-sorting policies; and (4) the promotion of waste-sorting policies affects knowledge dissemination. This study combines textual analysis and econometric techniques to offer a fresh perspective on the significance of public participation in promoting waste-sorting policies. It serves as an invaluable resource for governments to implement waste-sorting policies.

Heterodoxa dalam gerakan perlawanan perempuan Indonesia pada film “Kartini”

Patriarchal culture is a culture that is inherent and developing in Indonesia. This culture assumes a position of power in the male group. Kartini as an Indonesian hero is known to be able to produce ideas that are resistance in nature and has succeeded in influencing the culture of domination that exists to this day. Even though Kartini lived at a time when power relations were very strong, especially at a time when Indonesia was still not independent. Kartini's struggle against male domination was immortalized in the film entitled "Kartini". This research aims to see the process of Kartini's success in opposing culture and doxa through Pierre Bourdieu's thoughts regarding Post-Structuralism. In his thoughts, Bourdieu mentions several concepts, namely habitus, capital, arena and doxa which will then be associated with this film. This research uses descriptive qualitative methods with text analysis techniques to explain this process through existing film footage. From this research it can be concluded that Kartini has aspects of habitus, capital and arena that support the success of her thinking which originates from her position in society. So this can provide changes to the surrounding environment. Kartini's habitus, capital and arena are of great value and give her an advantage in establishing dominance in the area she is targeting. Heterodoxa requires high levels of dominance by its actors. The film "Kartini" shows how Kartini's dominance is able to shape a change, especially in the structures that previously existed in society.

Digital transformation and corporate green supply chain efficiency: Evidence from China

Enhanced green supply chain effectiveness is imperative to foster superior growth in manufacturing companies. In the era of the digital economy, businesses are increasingly embracing digital technology to drive sustainability. Therefore, this study uses data from Chinese manufacturing firms listed on the A-share market between 2011 and 2020. It employs text analysis techniques to derive indicators for digital transformation, and applies the DEA method to assess the efficiency of green supply chains while investigating their relationship. The findings indicate that digital transformation has significantly enhanced corporate green supply chain efficiency. The underlying mechanism suggests that digital transformation promotes innovation in green technology, reduces transaction costs, and alleviates financing constraints, thereby improving green supply chains’ overall efficiency. Moreover, the impact of digital transformation on enhancing efficiency in green supply chains is particularly pronounced among government-owned companies, large corporations, and those with lower per-capita production values. This study provides empirical evidence for firms seeking to achieve sustainable and green development and offers valuable policy insights for fostering superior economic growth.

Uncovering the quality factors driving the success of mobile payment apps

This study comprehensively analyzed the key quality factors of mobile payment applications that shape consumer perceptions and their influence on market performance. Leveraging text analytics techniques, including Latent Dirichlet Allocation (LDA) topic modeling and sentence-level sentiment analysis, we analyzed a vast dataset of 230,940 reviews of nine mobile payment apps. Our investigation revealed four significant quality factors discussed in online consumer reviews: payment convenience, security, customer service, and app design. Crucially, this study demonstrated the substantial impact of positive and negative evaluations of these factors on the market performance of mobile payment apps, with numerical ratings acting as mediators. This study breaks new ground by empirically validating the link between post-usage evaluations and the app success. Additionally, it also empirically validated the link between post-usage evaluations and app success, elucidating the role of quality factors. The insights can empower app developers and marketers to enhance user experience and attain market dominance.

Measuring trust: a text analysis approach to compare, contrast, and select trust questionnaires.

Trust has emerged as a prevalent construct to describe relationships between people and between people and technology in myriad domains. Across disciplines, researchers have relied on many different questionnaires to measure trust. The degree to which these questionnaires differ has not been systematically explored. In this paper, we use a word-embedding text analysis technique to identify the differences and common themes across the most used trust questionnaires and provide guidelines for questionnaire selection. A review was conducted to identify the existing trust questionnaires. In total, we included 46 trust questionnaires from three main domains (i.e., Automation, Humans, and E-commerce) with a total of 626 items measuring different trust layers (i.e., Dispositional, Learned, and Situational). Next, we encoded the words within each questionnaire using GloVe word embeddings and computed the embedding for each questionnaire item, and for each questionnaire. We reduced the dimensionality of the resulting dataset using UMAP to visualize these embeddings in scatterplots and implemented the visualization in a web app for interactive exploration of the questionnaires (https://areen.shinyapps.io/Trust_explorer/). At the word level, the semantic space serves to produce a lexicon of trust-related words. At the item and questionnaire level, the analysis provided recommendation on questionnaire selection based on the dispersion of questionnaires' items and at the domain and layer composition of each questionnaire. Along with the web app, the results help explore the semantic space of trust questionnaires and guide the questionnaire selection process. The results provide a novel means to compare and select trust questionnaires and to glean insights about trust from spoken dialog or written comments.

Investors' opinion disagreement and abnormal trading around pre-earnings announcements

We use textual analysis techniques to construct a direct measure of investor disagreement to explain the abnormal stock volume and return around earnings expectation announcement in China. We crawl scrape the most a popular Chinese stock forum for individuals' posts about their opinion on a stocks, and classify the posts with machine learning methods. We find that our measure is most correlated with return volatility and historical turnover. We then conduct an event study for abnormal stock volume around pre-earnings announcements, examining the cumulative abnormal turnover for sample groups with prior agreement or disagreement, and with a convergence or divergence of opinion. We find that given prior disagreement of opinion, both convergence and divergence of opinion causing more trading volume in the Chinese market. We also observe the sample group with a divergence of opinion has more cumulative abnormal return than the convergence group. However, this effect is found to be more profound in stocks with higher institutional ownership, which is believed to be correlated with a low short-sale constraint.

CEO overconfidence and financial reporting complexity: evidence from textual analysis

PurposeThis study investigates the relationship between the Chief Executive Officer's (CEO) overconfidence and financial reporting complexity in Iran, a context characterized by weak corporate governance and heightened managerial discretion.Design/methodology/approachThe sample consists of 1,445 firm-year observations from 2010 to 2021. CEO overconfidence (CEOOC) is evaluated using an investment-based index, specifically capital expenditures. Financial reporting complexity (Complexity) is measured through textual features, particularly three readability measures (Fog, SMOG and ARI) extracted from annual financial statements. The ordinary least squares (OLS) regression is employed to test the research hypothesis.FindingsResults suggest that CEOOC is positively related to Complexity, leading to reduced readability. Additionally, robustness analyses demonstrate that the relationship between CEOOC and Complexity is more distinct and significant for firms with lower profitability than those with higher profitability. This implies that overconfident CEOs in underperforming firms tend to increase complexity. Also, firms with better financial performance present a more positive tone in their annual financial statements, reflecting their superior performance. The findings remain robust to alternative measures of CEOOC and Complexity and are consistent after accounting for endogeneity issues using firm fixed-effects, propensity score matching (PSM), entropy balancing approach and instrumental variables method.Research limitations/implicationsThis study adds to the literature by delving into the effect of CEOs' overconfidence on financial reporting complexity, a facet not thoroughly investigated in prior studies. The paper pioneers the use of textual analysis techniques on Persian texts, marking a unique approach in financial reporting and a first for the Persian language. However, due to the inherent challenges of text mining and feature extraction, the results should be approached with caution.Practical implicationsThe insights from this study can guide investors in understanding the potential repercussions of CEOOC on financial reporting complexity. This will assist them in making informed investment decisions and monitoring the financial reporting practices of their invested companies. Policymakers and regulators can also reference this research when formulating policies to enhance financial reporting quality and ensure capital market transparency. The innovative application of textual analysis in this study might spur further research in other languages and contexts.Originality/valueThis research stands as the inaugural study to explore the relationship between CEOs' overconfidence and financial reporting complexity in both developed and developing capital markets. It thereby broadens the extant literature to include diverse capital market environments.

HistText: An Application for leveraging large-scale historical textbases

This paper introduces HistText, a pioneering tool devised to facilitate large-scale data mining in historical documents, specifically targeting Chinese sources. Developed in response to the challenges posed by the massive Modern China Textual Database, HistText emerges as a solution to efficiently extract and visualize valuable insights from billions of words spread across millions of documents. With a user-friendly interface, advanced text analysis techniques, and powerful data visualization capabilities, HistText offers a robust platform for digital humanities research. This paper explores the rationale behind HistText, underscores its key features, and provides a comprehensive guide for its effective utilization, thus highlighting its potential to substantially enhance the realm of computational humanities.

Building care gerontotechnologies in the context of elderly person with Alzheimer's disease

Objective: to build care gerontotechnologies to help family caregivers with the difficulties experienced in the context of elderly people with Alzheimer's disease. Methods: strategic action research carried out with seven family caregivers of elderly people with Alzheimer's disease and 12 health academics from a university. Data was gathered from the family caregivers using a semi-structured interview. Three focus groups were held with the academics. The data was submitted to the discursive textual analysis technique. Results: eight difficulties experienced by family caregivers emerged, corresponding to cognitive aspects, Activities of Daily Living and issues relating to family care. It was possible to construct 14 care gerontotechnologies in product form. Conclusion: gerontotechnologies were built to help family caregivers with the difficulties experienced by elderly people living with Alzheimer's disease. Contributions to practice: the gerontotechnologies developed have the potential to contribute to the care process, since they can be used by family caregivers on a daily basic and can be adapted to each reality, based on the unique needs of each elderly person.