Text Mining Methods Research Articles

공시연구의 네트워크 중심성 분석

[Purpose] This study aimed to examine the direction of domestic research through text analysis using a qualitative method rather than an empirical study in the name of ‘public disclosure’, targeting domestic research that is mostly empirical research. [Methodology] This study conducted frequency analysis on the research sample, focusing on keywords from a total of 872 accounting－related public disclosure studies, and conducted four types of centrality and network analysis based on the frequency analysis. For this analysis, R and Python programs were used. [Findings] The results of this study are as follows. First, among a total of 872 domestic studies, the keywords with the highest frequency were accounting, profit, and company in that order. Second, as a result of the Degree centrality analysis, the key keywords were profit (0.0736), company (0.0664), and information (0.0564). Third, as a result of Closeness centrality analysis, the keywords showing connectivity that emphasized the distance between one node and all nodes in the entire network were profit (0.3468), company (0.3392), and information (0.3349). Fourth, as a result of Betweenness centrality analysis, the keywords shown to be most helpful in building a network of other nodes are profit (0.1499), company (0.1172), accounting (0.0990), and information (0.0842). Fifth, as a result of the Eigenvector centrality analysis, as an indicator that reflects the centrality of nodes in the weight, the keywords with the highest centrality were accounting (0.6136), international accounting standards (0.5371), and standards (0.3988). [Implications] Through this study, which is not an empirical study but a qualitative analysis using a text mining method, it is expected that later researchers will have an easier time analyzing research trends in previous studies, and furthermore, it is expected that further expansion of research topics will be possible.

Assessing Data Quality in the Age of Digital Social Research: A Systematic Review

While survey data has long been the focus of quantitative social science analyses, observational and content data, although long-established, are gaining renewed attention; especially when this type of data is obtained by and for observing digital content and behavior. Today, digital technologies allow social scientists to track “everyday behavior” and to extract opinions from public discussions on online platforms. These new types of digital traces of human behavior, together with computational methods for analyzing them, have opened new avenues for analyzing, understanding, and addressing social science research questions. However, even the most innovative and extensive amounts of data are hollow if they are not of high quality. But what does data quality mean for modern social science data? To investigate this rather abstract question the present study focuses on four objectives. First, we provide researchers with a decision tree to identify appropriate data quality frameworks for a given use case. Second, we determine which data types and quality dimensions are already addressed in the existing frameworks. Third, we identify gaps with respect to different data types and data quality dimensions within the existing frameworks which need to be filled. And fourth, we provide a detailed literature overview for the intrinsic and extrinsic perspectives on data quality. By conducting a systematic literature review based on text mining methods, we identified and reviewed 58 data quality frameworks. In our decision tree, the three categories, namely, data type, the perspective it takes, and its level of granularity, help researchers to find appropriate data quality frameworks. We, furthermore, discovered gaps in the available frameworks with respect to visual and especially linked data and point out in our review that even famous frameworks might miss important aspects. The article ends with a critical discussion of the current state of the literature and potential future research avenues.

Anti-Semitic Speech Detection and Classification in Online Social Network using Deep Learning

Every individual possesses the entitlement to freedom of speech. However, in the guise of free expression, this privilege is being abused to discriminate against and harm other people. This prejudice is referred to as hate speech. A clear definition of hate speech is language that expresses hatred for an individual or a group of individuals based on traits including race, religion, ethnicity, gender, nationality, handicap, and sexual orientation. Hate speech has become increasingly widespread, both in physical spaces and on the internet, in recent years. Thus, recent studies used a range of machine learning and deep learning techniques with text mining method to automatically recognise the hate speech messages on real-time datasets in order to address this developing issue in social media sites. This project's goal is to examine comments on social networks using Natural Language Processing (NLP) and a Deep Learning method called VADER method. In order to identify the text as positive or negative, VADER neural networks are used to extract the keywords from user generated content. If it's negative, immediately block the comments in accordance with the user's preferences and block the friends in accordance with pre-established threshold values. The proposed framework was deployed in a real-time social networking site with an improved notification system, according to experimental findings

Relationship between quality management and organizational performance in the healthcare industry

ABSTRACT This study investigates how Quality Management (QM) programmes, particularly the Baldrige excellence framework, provide an approach to enhance healthcare organisational performance. Two independent methods were used to examine the short-term and long-term effects of winning the Baldrige award and the specific influence of common QM practices on results. The first method assesses the Baldrige Award’s impact on hospital performance by examining patient survey data and shows that initial quality improvements are not sustained in the long term. The second method uses confirmatory semantic analysis (CSA), a text-mining method, to analyse 22 Baldrige award-winning applications to determine the relationship between QM practices and organisational results in healthcare settings. The results show that only three QM practices – measurement, analysis, and knowledge management (MAKM), leadership, and operations focus – significantly contribute to organisational performance. Furthermore, CSA findings highlight the dominant role of leadership in driving performance outcomes, with other factors in the framework showing nonsignificant mediation effects. By exploring both the overarching effectiveness of QM programmes across different time lengths and specific practices that contribute to performance, this research provides academic insights with practical application in the healthcare industry. Importantly, the findings highlight the nuanced relationship between QM initiatives and organisational excellence.

Exploring cybertechnology standards through bibliometrics: Case of National Institute of Standards and Technology

Cyber security is one of the topics that gain importance today. It is necessary to determine the basic components, basic dynamics, and main actors of the Cyber security issue, which is obvious that it will have an impact in many areas from social, social, economic, environmental, and political aspects, as a hot research topic. When the subject literature is examined, it has become a trend-forming research subject followed by institutions and organizations that produce R&D policy, starting from the level of governments. In this study, cybersecurity research is examined in the context of 5 basic cyber security functions specified in the cyber security standard (CSF) defined by the National Institute of Standards and Technology (NIST). It is aimed to determine the research topics emerging in the international literature, to identify the most productive countries, to determine the rankings created by these countries according to their functions, to determine the research clusters and research focuses. In the study, several quantitative methods were used, especially scientometrics, social network analysis (SNA) line theory and structural hole analysis. Statistical tests (Log-Likelihood Ratio) were used to reveal the prominent areas, and the text mining method was also used. we first defined a workflow according to the “Identify”, “Protect”, “Detect”, “Respond” and “Recover” setups, and conducted an online search on the Web of Science (WoS) to access the information on the publications on the relevant topics It is seen that actors, institutions and research create different densities according to various geographical regions in the 5 functions defined within the framework of cybersecurity. It is possible to say that infiltration detection, the internet of things and the concept of artificial intelligence are among the other prominent research focuses, although it is seen that smart grids are among the most prominent research topics. In the first clustering analysis we performed, we can say that 17 clusters are formed, especially when we look under the definition function. The largest of these clusters has 32 data points, so-called "decision making models".

Research on the Sentiment Analysis and Evolutionary Mechanism of Sudden Network Public Opinion Based on SNA-ARIMA Model with Text Mining

This article is based on the massive text data in the digital era, using text mining methods to explore the evolution mechanism of online public opinion, in order to achieve the control of public opinion trends in sudden network events and regulate the stability of online and offline public opinion. We introduced time series for decision-making assistance, combined with time series autoregressive integrated moving average model (ARIMA) for prediction, and comprehensively evaluates the evolution of public opinion from two dimensions: sentimental orientation and social network analysis (SNA). Based on this, a new exploration model for public opinion evolution, SNA-ARIMA, is constructed. The research results indicate that the process of public opinion dissemination exhibits distinct phased characteristics, and each stage is comprehensively influenced by different key nodes or factors. This study provides decision support for managing public opinion crises, making the identification of key nodes in online public opinion events, the evolution mechanism of public opinion, and the guidance of public opinion crises more systematic, forward-looking, and scientific.

Natural language processing (NLP) to facilitate abstract review in medical research: the application of BioBERT to exploring the 20-year use of NLP in medical research

Background review is a time and labor-consuming step in the systematic and scoping literature review in medicine. Text mining methods, typically natural language processing (NLP), may efficiently replace manual abstract screening. This study applies NLP to a deliberately selected literature review problem, the trend of using NLP in medical research, to demonstrate the performance of this automated abstract review model.MethodsScanning PubMed, Embase, PsycINFO, and CINAHL databases, we identified 22,294 with a final selection of 12,817 English abstracts published between 2000 and 2021. We invented a manual classification of medical fields, three variables, i.e., the context of use (COU), text source (TS), and primary research field (PRF). A training dataset was developed after reviewing 485 abstracts. We used a language model called Bidirectional Encoder Representations from Transformers to classify the abstracts. To evaluate the performance of the trained models, we report a micro f1-score and accuracy.Results The trained models’ micro f1-score for classifying abstracts, into three variables were 77.35% for COU, 76.24% for TS, and 85.64% for PRF.The average annual growth rate (AAGR) of the publications was 20.99% between 2000 and 2020 (72.01 articles (95% CI: 56.80–78.30) yearly increase), with 81.76% of the abstracts published between 2010 and 2020. Studies on neoplasms constituted 27.66% of the entire corpus with an AAGR of 42.41%, followed by studies on mental conditions (AAGR = 39.28%). While electronic health or medical records comprised the highest proportion of text sources (57.12%), omics databases had the highest growth among all text sources with an AAGR of 65.08%. The most common NLP application was clinical decision support (25.45%).ConclusionsBioBERT showed an acceptable performance in the abstract review. If future research shows the high performance of this language model, it can reliably replace manual abstract reviews.

Effects of robots’ appearance on guest service experiences

ABSTRACT Robots are starting to replace human employees in the accommodation industry. It remains unclear how guests perceive robots. The study aims to evaluate the effectiveness of the humanoid and non-humanoid robots and to compare the guests’ perceptions. The data was collected by web digging and analyzed using text mining and machine learning. The results show that guests experience more positive emotions in non-humanoid robotic experiences. They also stated that they found both types of robot fun and interesting. However, guests have negative perceptions about the perceived usefulness and appearance of humanoid robots. According to the findings of the machine learning analysis, the neural network and logistic regression models show the best statistical performance in both datasets compared to other machine learning models. The study contributes to the literature through an investigation of guests’ opinions regarding robot types with the use of machine learning and text mining methods.

Promoting Sustainable Development in Urban–Rural Areas: A New Approach for Evaluating the Policies of Characteristic Towns in China

The Sustainable Development Goals, adopted by all United Nations Member States, emphasize sustainable cities and communities, aiming to strengthen development planning and foster positive links between urban and rural areas by 2030. As the newly leading platform for urbanization development, characteristic towns (‘Tese Xiaozhen’ in Chinese) contribute to the target of sustainable urbanisation, wherein town policies are especially important for leading the sound development of characteristic towns. However, the effect of these characteristic towns policies remains uncertain. Additionally, most related studies have utilized qualitative methodologies in policy evaluation, which may be inadequate to guide actual practice. Thus, in order to accurately acknowledge the effect of characteristic towns policies, a policy evaluation framework is established in this paper, utilising the multiple streams theory, text mining and the Policy Modelling Consistency Index model (PMC-Index model). Then, taking 225 policy documents promulgated in China from 2015 to 2022 as the research objects, policy evaluation indicators were selected by combining the multiple streams theory to improve the PMC-Index model and a text mining method. Thereafter, an empirical analysis was conducted to evaluate the consistency of 10 characteristic towns policies chosen. The results indicate that the mean value of the PMC indexes of the 10 examined policies is 7.13 in the total of 9 points, which means that the general performance of the characteristic town policies is satisfactory, yet the low scores in terms of effectiveness, content, and guarantee. In the end, some recommendations and suggestions are proposed for clarifying administrative attributes, optimizing the integration of urban and rural resources, and adjusting the policy hierarchy to a bottom-up approach. This study not only provides a policy evaluation framework to comprehensively understand the consistency of global urbanization policies but also offers a beneficial reference for promote urban–rural sustainable development.

Severity Analysis for Occupational Heat-related Injury Using the Multinomial Logit Model

BackgroundWorkers are often exposed to hazardous heat due to their work environment, leading to various injuries. As a result of climate change, heat-related injuries (HRIs) are becoming more problematic. This study aims to identify critical contributing factors to the severity of occupational HRIs. MethodsThis study analyzed historical injury reports from the Occupational Safety and Health Administration (OSHA). Contributing factors to the severity of HRIs were identified using text mining and model-free machine learning methods. The Multinomial Logit Model (MNL) was applied to explore the relationship between impact factors and the severity of HRIs. ResultsThe results indicated a higher risk of fatal HRIs among middle-aged, older, and male workers, particularly in the construction, service, manufacturing, and agriculture industries. In addition, a higher heat index, collapses, heart attacks, and fall accidents increased the severity of HRIs, while symptoms such as dehydration, dizziness, cramps, faintness, and vomiting reduced the likelihood of fatal HRIs. ConclusionsThe severity of HRIs was significantly influenced by factors like workers’ age, gender, industry type, heat index , symptoms, and secondary injuries. The findings underscore the need for tailored preventive strategies and training across different worker groups to mitigate HRIs risks.

A systematic evaluation of text mining methods for short texts: Mapping individuals’ internal states from online posts

Short texts generated by individuals in online environments can provide social and behavioral scientists with rich insights into these individuals’ internal states. Trained manual coders can reliably interpret expressions of such internal states in text. However, manual coding imposes restrictions on the number of texts that can be analyzed, limiting our ability to extract insights from large-scale textual data. We evaluate the performance of several automatic text analysis methods in approximating trained human coders’ evaluations across four coding tasks encompassing expressions of motives, norms, emotions, and stances. Our findings suggest that commonly used dictionaries, although performing well in identifying infrequent categories, generate false positives too frequently compared to other methods. We show that large language models trained on manually coded data yield the highest performance across all case studies. However, there are also instances where simpler methods show almost equal performance. Additionally, we evaluate the effectiveness of cutting-edge generative language models like GPT-4 in coding texts for internal states with the help of short instructions (so-called zero-shot classification). While promising, these models fall short of the performance of models trained on manually analyzed data. We discuss the strengths and weaknesses of various models and explore the trade-offs between model complexity and performance in different applications. Our work informs social and behavioral scientists of the challenges associated with text mining of large textual datasets, while providing best-practice recommendations.

Clustering Analysis of Hotel Network Reviews Based on Text Mining Method

With the development of information technology, users use online platforms to post real-time online comments to express their preferences and opinions on goods or services. Online review information expresses users' behavioral habits and special preferences. In depth, analysis of hotel online reviews can improve the adaptability of hotel services to user needs. Effective mining of the vast user review data will provide value for the development of the tourism industry. Using text mining methods to process hotel review data, multiple clustering methods were compared and analyzed for positive and negative feature words from the perspective of user experience. It was found that the k-means++ algorithm had a better clustering effect on user network reviews and achieved better clustering and segmentation of user evaluation information. Unsupervised clustering analysis can be used to further classify online comment information into categories based on positive and negative reviews, providing intellectual support for improving the precision and personalized service quality of hotels.

텍스트 마이닝을 이용한 코로나19 전후 청소년과 식생활 관련 키워드 검색 경향 분석

This study analyzed Naver, Daum, Google, YouTube, and Twitter using TEXTOM for two years and four years as of January 18, 2020. The results are as follows. First, the total number and volume of keyword search data related to youth and diet were slightly higher after COVID-19, showing that interest increased due to COVID-19. Second, as a result of frequency analysis, ‘education’ was the highest before COVID-19, and ‘health’ was the highest after COVID-19, showing that interest in health is increasing due to the increased importance of health and immunity due to COVID-19. Third, as a result of frequency weight analysis of the top 50 keywords, ‘education’ showed the highest frequency before COVID-19, and ‘acne’ after COVID-19. Fourth, the results visualized using word cloud showed that the keywords ‘education’ before COVID-19 and ‘health’ after COVID-19 appeared the largest and boldest, showing the highest frequency and importance. As a result of the above results, we were able to use the text mining method to apply it to eating habits, and we used materials visualized as a highly readable word cloud in units such as eating problems in adolescence and balanced meal planning and selection in the home economics curriculum to improve the teaching of the class. The direction of proper eating habits education, including using it as a medium, was presented.

Robotic Solutions for the Challenges of Human Labor: Managers Perspective

Managers have to endure some costs in order to develop human resources, which is one of the most significant providers of organizational performance. These costs are expenditures for human resources activities and customer dissatisfaction caused by work accidents and negative personnel behaviors. This study primarily aims to determine the expenditures made by the managers and the behavior of the employees that disturb the guests. Text mining methods is used to identify cost of human labor and employee behaviors referred to by managers' reviews, and association rules is used to find out common cost and behaviors in managers' reviews. The secondary aim of the study is to investigate the managers' intentions to employ robots to solve human-induced challenges. Contrary to the literature, managers believe that robots cannot contribute to the solution of these problems and they don’t think to work with service robots.

Towards a ‘synergy’ of text mining and critical discourse analysis: a corpus-assisted discourse study of imagining China in Hong Kong political discourse

Abstract This study proposes to incorporate text mining into critical discourse analysis (CDA) to give a corpus-assisted discourse study of the particular ways of imagining Hong Kong’s relations to China in the public speeches of three former Chief Executives in the two decades after its handover. With the computer-assisted text-mining tool KH Coder, this study combines the methods of quantitative text mining and qualitative discourse analysis to examine their preferential ways of imagining China at different levels of discourse: (1) topics/themes, (2) discursive strategies, and (3) linguistic means and realizations. It generates illuminating findings concerning their preferential ways of imagining Hong Kong’s relations to China as well as the analytic potential of incorporating text mining into CDA. It is expected that it can lead to more studies towards this endeavour.

Can local fintech development improve analysts’ earnings forecast accuracy? Evidence from China

This paper examines the impact of local fintech development on the accuracy of analysts’ earnings forecasts. We employ the method of web text mining to construct the local fintech development index for empirical test. Our findings indicate that local fintech development has a substantial positive impact on analysts’ earnings forecast accuracy by promoting firm digital transformation and alleviating the information asymmetry between firms and outsiders. Additionally, this effect is particularly pronounced for analysts without equity pledge affiliation and those with weaker buy-side pressure. This study shows that local fintech development can enhance the information environment of the capital market.

Sentiment Analysis of Indonesian Youtube Reviews About Lesbian, Guy, Bisexual and Transgender (LGBT) using IndoBERT Fine Tuning

Sentiment Analysis of Indonesian Youtube Reviews About Lesbian, Guy, Bisexual and Transgender (LGBT) using IndoBERT Fine Tuning

Text Mining technologies in sociological analysis (using the example of studying students`ideas about the mission of a modern university)

There are discussed in the article the possibilities of using Text Mining methods in the practice of analyzing the information received on the base of open questionnaire questions. The paper presents an example of unigrams and bigrams analysis, as well as the search for latent topic using thematic modeling. Empirical materials present the data of survey conducted in 2022, in which 929 students of one Moscow economics university took part. In the open question of the questionnaire, it was proposed to define the mission of the university. Information made it possible to get the subjective interpretation of the main significancy of higher education in modern conditions. The frequency analysis of unigrams, supplemented by a qualitative analysis of respondents’ statements, allowed reflecting the vocabulary of student discourse about the mission of the university. The articulation of bigrams was carried out on the basis of several statistical metrics, which made it possible to rank phrases and highlight a key set of concepts. The procedure revealed that in the perception of students, the priorities of the university are aimed at the transferring of professional knowledge and skills, in a broad sense – the training of qualified specialists. The social functions of the university, focused on meeting the needs of society and the state, are less pronounced in the conceptual interpretations of the interviewed students. At the next stage of the study the task of articulation and research of latent topics was put forward. The specific feature of thematic modeling is that the words combined into one topic reflect the distribution of words identified by the program, but not a topic that is literally understandable to a person. Taking into account the specifics of the method used, the authors demonstrated the results of search analysis in the practice of processing an open question. As it turned out, the keywords concentrated in the core of the main topics are mainly related to meeting the needs of the students themselves, leaving on the periphery of the verbalized definitions any understanding of the importance of the university as a platform for innovation, scientific research, entrepreneurial and other initiatives for the benefit of society and the country. The results of the presented research can be useful in rethinking the research tools of sociologists in the context of the active development of digital technologies, which requires testing new methods, understanding their real capabilities and limitations in solving the tasks of sociological research.

INTRICATE GENEALOGIES: WHAT IS SAID IN THE EPIC POEM ABOUT NOGAY PEOPLE?

This article examines communication blueprints of the Forty Knights of Steppe epic poem (original title in Kazakh "Қырымның қырық батыры"). The Forty Knights of Steppe is a cultural heritage of Turkic nations. The epic poem’s text is an assemblage of ballads that are connected with each other. It consists of thirty-five ballads and depicts relationships of around two hundred characters. The epic poem's events have a historical origin that links to the Nogay nobility from the Golden Horde period. The main event in the epic poem develops around Edige and his descendants. The historical aspect of this epic poem positions it with the paragons of heroic literature such as "Shahnameh" and "Manas". The epic poem’s text was first fully recorded in 1942 in Almaty by Muryn Zhyrau (real name ‘Tilegen Sengirbekuly’) by specialists from the Kazakh SSR Academy of Sciences. Unfortunately, the epic poem stayed unpublished for more than 60 years. The full text of the epic poem was published in 2005 after restoring the independence of the Republic of Kazakhstan. This research discovers the epic poem by analyzing their network. It focuses on mapping and visualizing links between characters described in the text. The authors used the methods of text mining, proposition, and semantic triangles for assembling information about contacts between characters into the databases. The databases include basic information, such as the names of characters and their contacts with other characters. The databases also contain information about the direction of every contact, relationship statuses of contactors, contactors' roles inside of genealogical lineage, kinship degrees of contactors inside their genealogies, and contactors' gender. In addition, the databases include information about the connection of contactors to genealogies, and information about the contacts of contactors outside of their genealogies. All databases were built in spatial data analysis compatible format. The databases were launched through data visualization software by switching on the environmental settings required for this type of research. The authors used Gephi data visualization software for this research to visualize databases. The visualization shows an intricate communicative network that covers almost all characters. The visualization demonstrates two types of contacts between characters of the epic poem. The first type is a structured type of contacts. This type of contacts develops inside of genealogies. It is a common type of contacts in the epic poem, and it follows the hierarchical order. The second type is a class-based type of contacts. This type of contacts ties characters from different genealogies and does not correspond to the hierarchical order. However, this type of contacts depends on the characters’ roles and statuses inside their genealogies. Thereby, the second type of contacts is connecting different genealogies into whole network. The types of contacts demonstrate two levels of social communication in the narration of the epic poem. Considering the visualization results and according to the historical origins of the characters, the authors argue that the text of the epic poem shows the patterns of social communication typical for Central Eurasian medieval nomad cultures in the Golden Horde period. On the other side, the stratification into two types of contacts may demonstrate the levels of bureaucracy in social communication in the same historical period. The authors suggest that the social networks of the epic poem describe the transitional form of the chiefdom society with communication levels typical for tribal and chiefdom societies. In general, the paper's authors suppose that the example of this research on the epic poem’s communication structure may give more data to understand the correlation between language and society.

Research on Barrage Text Mining and Emotional Orientation in Online Interactive Learning Videos: Based on the Case Study of "Luo Xiang's Criminal Law"

This study uses text mining and sentiment analysis methods to analyze the barrage text information of "Luo Xiang's Criminal Law" online teaching videos and sort out the main emotional tendencies of users. The research found that user emotions change with the dispersion of hot topics, showing a decentralized characteristic, and at the same time, users still have a certain lack of capture for the content of online teaching videos in the interactive learning process. It is urgent to clarify the specific impact of technology on users. This study is helpful for deepening research on online interactive learning, clarifying user emotional tendencies, and providing a decision-making basis for accurately improving interactive learning effects.