Text Network Analysis Research Articles

A study on research trends for pregnancy in adolescence: Focusing on text network analysis and topic modeling

A study on research trends for pregnancy in adolescence: Focusing on text network analysis and topic modeling

Mapping the terrain of hate: identifying and analyzing online communities and political parties engaged in hate speech against Muslims and LGBTQ+ communities

AbstractThis study investigates the impact of X on political discourse and hate speech in Finland, focusing on Muslim and LGBTQ+ communities from 2018 to 2023. During this period, these groups have experienced increased hate speech and a concerning surge in hate crimes. Utilizing network analysis methods, we identified online communities and examined the interactions between Finnish MPs and these communities. Our investigation centered on uncovering the emergence of networks propagating hate speech, assessing the involvement of political figures, and exploring the formation dynamics of digital communities. Employing agenda-setting theory and methodologies including text classification, topic modeling, network analysis, and correspondence analysis, the research uncovers varied communication patterns in retweet and mention networks. Retweet networks tend to be more fragmented and smaller, with participation primarily from far-right Finns Party MPs, whereas mention networks exhibit wider political representation, including members from all parties. Findings highlight the Finns Party MPs' significant role in fostering divisive, emotionally charged communications within politically segregated retweet communities, contrasting with their broader engagement in mention networks. The study underscores the necessity for cross-party efforts to combat hate speech, promote inclusive dialogue, and mitigate political polarization.

Quantifying the synergy of China's carbon neutrality policies through policy documents

Carbon neutrality is a crucial pathway to addressing global climate challenges. To achieve the goal of carbon neutrality by 2060, the Chinese government has introduced a series of policies at various levels. Synergizing between these different levels of government and policy domains is crucial for smoothly advancing carbon neutrality goals. Additionally, policy documents provide a lens through which to observe policy synergy. This study defines a framework for assessing the synergy of carbon neutrality policies from the perspectives of policy content, policy issuer, and the temporal dimension. Utilizing text mining techniques and network analysis methods, the study quantifies policy synergy based on 224 carbon neutrality policies released by the Chinese government. The findings reveal high content synergy for the policy tool and goal dimensions, with network densities of 0.78 and 0.71, respectively. Regarding issuer synergy, the synergy across different regions in China varies significantly, ranging from 0.653 to 0.912. Meanwhile, as policy topics evolved, the temporal synergy increased from 0.88 to 0.90. These conclusions provide robust support for policy formulation, offering valuable insights to realize global carbon neutrality goals more efficiently and comprehensively.

A Study on MBTI Perceptions in South Korea: Big Data Analysis from the Perspective of Applying MBTI to Contribute to the Sustainable Growth of Communities

This study aimed to assess the potential contributions of the Myers–Briggs Type Indicator (MBTI) to the sustainable growth of communities by conducting a comprehensive analysis of social perceptions of the MBTI in South Korea through big data analysis. The investigation encompasses three primary stages: data collection, preprocessing, and analysis, involving text mining, network analysis, CONCOR analysis, and sentiment analysis. A total of 31,308 text data pieces (13.73 MB) from various sources, including news, blogs, and other sections of Naver and Google, over the past three years, were collected and analyzed using the keyword “MBTI”. Tools, such as Textom SV, UCINET, and NetDraw, were employed for data collection and analysis. The study’s key findings include the identification, through term frequency (TF) and TF-inverse document frequency analyses, of top-ranking terms, such as 16Types, 4Indicators, Test, Myself, OthersMBTI, Situation, and Contents. The CONCOR analysis further revealed six clusters, encompassing themes like interest in MBTI personality tests, application of 16 types in daily life, MZ’s MBTI consumption patterns, trending of MBTI characters, extension to K-Test, and professional use of MBTI. Moreover, sentiment analysis indicated that 68.5% of individuals in South Korea expressed a positive sentiment towards MBTI, while 31.5% conveyed a negative sentiment. The specific emotions identified included liking (Good Feeling), disgust, and interest, in order of prominence. In light of these findings, this study delineates a spectrum of perceptions regarding MBTI in South Korea, encompassing both positive interests and negative concerns. To ensure the responsible use of MBTI, it is imperative to implement reliable scientific testing and education, mitigate the potential harm of stereotyping, and reshape social perceptions surrounding MBTI usage. Only through these measures can MBTI genuinely contribute to the sustainable growth of communities without being confined to limiting stereotypes.

Scaling up Archival Text Analysis with the Blockmodeling of N-gram Networks: A Case Study of Bulgaria’s Representation in the “Osservatore Romanoˮ (January – May 1877)

This paper seeks to bridge the gap between archival text analysis and network analysis by applying network clustering methods to analyze the coverage of Bulgaria in 123 issues of the newspaper Osservatore Romano published between January and May 1877. Utilizing optical character recognition and generalized homogeneity blockmodeling, the study constructs networks of relevant keywords. Those including the sets Bulgaria and Russia are rather isomorphic and they largely overlap with those for Germany, Britain, and War. In structural terms, the blockmodel of the two networks exhibits a clear core-semiperiphery-periphery structure that reflects relations between concepts in the newpaper’s coverage. The newspaper’s lexical choices effectively delegitimised the Bulgarian national revival, highlighting the influence of the Holy See on the newspaper’s editorial line.

Trend Research on Maritime Autonomous Surface Ships (MASSs) Based on Shipboard Electronics: Focusing on Text Mining and Network Analysis

The growing adoption of electric propulsion systems in Maritime Autonomous Surface Ships (MASSs) necessitates advancements in shipboard electronics for safe, efficient, and reliable operation. These advancements are crucial for tasks such as real-time sensor data processing, control algorithms for autonomous navigation, and robust decision-making capabilities. This study investigates research trends in MASSs, using bibliographic analysis to identify policy and future research directions in this evolving field. We analyze 3363 MASS-related articles from the Web of Science database, employing co-occurrence word analysis and latent Dirichlet allocation (LDA) topic modeling. The findings reveal a rapidly growing field dominated by image recognition research. Keywords such as “datum”, “image”, and “detection” suggest a focus on collecting and analyzing marine data, particularly with deep learning for synthetic aperture radar imagery. LDA confirms this, with “image analysis and classification research” as the leading topic. The study also identifies national and organizational leaders in MASS research. However, research on Arctic routes lags behind that on other areas. This work provides valuable insights for policymakers and researchers, promoting a deeper understanding of MASSs and informing future policy and research agendas regarding the integration of electric propulsion systems within the maritime industry.

Exploring over a decade of systems engineering research center: A community detection and text analytics approach

AbstractThe Systems Engineering Research Center (SERC) is a University Affiliated Research Center (UARC) of the US Department of Defense (DoD) formed in 2008 with more than 20 collaborator universities in the United States. Over the last decade, SERC has conducted research with Principal Investigators from universities within the SERC network, as reflected in technical reports (TR). These reports describe detailed information and analysis of the conducted research for every project under SERC support, such as written records of experiments or results of a scientific project. We analyzed the TRs from 2009 to early 2023 to identify research streams, topics, and evolution in systems engineering (SE) research using text mining and network analysis techniques, such as Louvain Community Detection and word similarity. As a result, we identified four major research streams over a decade of research projects, along with insights about topics and the evolution of SE within this time frame. Finally, we distinguished most profile authors and their most significant collaborations and networks.

‘Wolf Warriors’ in the UN Security Council? Investigating power shifts through blaming

AbstractWhat shifts in the extent and patterns of China's blaming are observable at the United Nations Security Council (UNSC)? This article employs automated content analysis to measure the frequency of blaming by China and other permanent UNSC member states based on a dataset of all speeches (19,623 overall) in the UNSC from January 1995 to May 2022. Furthermore, it uses qualitative text analysis and network analysis to examine the targets of blaming in three different case studies of UN security governance—UN peacekeeping in former Yugoslavia (1998–2008), the UN sanctions regime against North Korea (2007–2017) and the broader peace and security discourse at the UNSC (2008–2018). The study conceptualizes Chinese blaming as a previously unexamined discursive strategy of compulsory power with potential long‐term effects on shifts in productive power. However, the results indicate the absence of blaming‐related shifts in the discursive space of the UNSC: China's rhetoric in the UNSC, compared with other UNSC members, has remained restrained over time. Variance in blaming behaviour between actors and across bilateral and multilateral settings seems to depend on strategic considerations that assess how to exert power most effectively.

학교상담자가 인식하는 ‘좋은 학교상담자’의 특성

The purpose of this study is to understand the perception of ‘good school-counselors’ in school counseling. For this purpose, data were collected through open-ended questions about 'good school counselors' from 146 school counselors who agreed to participate in the study. Text network analysis method was used to analyze the collected data. As a result of the study, the most central words in school counselors' perception of a 'good school counselor' were 'client', 'counseling', and 'understanding'. As a result of analyzing the substructure of 'good school counselors', five communities emerged. The five communities were named ‘self-improvement as an expert’, ‘systematic application ability’, ‘ability to solve clients' problem’, excellent collaboration ability as a school member’, ‘student centered and permissive counseling attitudes’, and ‘attitude to realize a positive influence through counseling’. In this regard, the discussion revealed the implications of the research results and the limitations of the research.

Construction and Evaluation of a Traditional Village Tourism Imagery Element System Based on Xidi Village

Tourism imagery is one of the decision variables for tourists’ travel intentions and behavior. Traditional village tourism imagery studies mostly focus on tourist space, landscape imagery, etc. Few overall imagery studies have been conducted. This paper integrates non-structural and structural methods, uses three-level coding based on rooting theory, constructs a traditional village tourism place imagery element system through network text analysis, designs and analyzes questionnaires accordingly, and empirically evaluates the tourism imagery of Xidi village in Anhui Province. The results show that the overall satisfaction of tourists is less than important, indicating that there is a systematic deficiency in the perception of tourism place imagery of Xidi village. The dimensional IPA analysis shows that the three dimensions of natural landscape, social landscape, and physical cuisine are in the improvement area. According to the supplementary IPA factor analysis, factors such as natural phenomena, historical figures, clan history, watershed, accommodation, ticket systems, and guide services need important attention, improvement, and promotion. The results of the emotional analysis are dominated by positive images, but attention needs to be paid to the negative impact of negative images. Based on the above findings, recommendations are proposed to improve the layout of the development, and enhance the service facilities and quality based on these findings.

텍스트 마이닝을 활용한 종교교육학 연구동향 분석: 불교교육학과 기독교교육학을 중심으로

[Objective] The purpose of this study is to analyze research trends in Buddhist and Christian education using text mining techniques.
 [Contents] This study applied Python text mining techniques to automate much of the research process and utilized ChatGPT's API for tasks such as summarizing papers and extracting topics. Academic papers and theses were automatically collected from riss.kr using keywords ‘Buddhist education’ and ‘Christian education’, followed by preprocessing to remove duplicate and irrelevant papers. Trends in research by year and academic society were visualized and analyzed. Through processes such as morphological analysis, noun extraction, and paper summarization, the study identified research trends using keyword analysis, text network analysis, and topic modeling techniques.
 [Conclusion] Research in Christian education was approximately ten times more abundant than in Buddhist education. While research in Christian education is conducted diversely by the KSCRE and various Christian universities, Buddhist education is mainly led by the KASRE and Dongguk University. Keyword analysis and topic modeling revealed that research in Buddhist education focuses on philosophical and abstract topics such as intellectual practice, meditation, ethics, and death, whereas research in Christian education tends to focus on concrete and practical dimensions such as faith in the family, educational curriculum, and career guidance.

Interpretation of Image Symbols in Tourist Destinations based on Network Text Analysis: A Case Study of Lu'an

A good image of a tourist destination is the core competitiveness and a strong factor to attract tourists. The large use of new media in life and the popularization of the Internet make the online text gradually become an important means for tourists to obtain the tourist destination information. Taking Lu 'an as an example, based on the online travel notes collected by Ctrip, Qunar and Hornet' s Nest tourism websites, this paper uses ROST Content Mining6 to perceive the tourist destination, and found that tourists have a strong perception of natural tourism resources and low experience of transportation, accommodation and other tourism reception facilities. Lu 'an should develop targeted tourism products with local characteristics, improve tourism service facilities, design diversified tourism routes to improve tourists' satisfaction and improve the image of a tourist destination.

Big data meets storytelling: using machine learning to predict popular fanfiction

Fanfictions are a popular literature genre in which writers reuse a universe, for example to transform heteronormative relationships with queer characters or to bring romance into shows focused on horror and adventure. Fanfictions have been the subject of numerous studies in text mining and network analysis, which used Natural Language Processing (NLP) techniques to compare fanfictions with the original scripts or to make various predictions. In this paper, we use NLP to predict the popularity of a story and examine which features contribute to popularity. This endeavor is important given the rising use of AI assistants and the ongoing interest in generating text with desirable characteristics. We used the main two websites to collect fan stories (Fanfiction.net and Archives Of Our Own) on Supernatural, which has been the subject of numerous scholarly works. We extracted high-level features such as the main character and sentiments from 79,288 of these stories and used the features in a binary classification supported by tree-based methods, ensemble methods (random forest), neural networks, and Support Vector Machines. Our optimized classifiers correctly identified popular stories in four out of five cases. By relating features to classification outcomes using SHAP values, we found that fans prefer longer stories with a wider vocabulary, which can inform the prompts of AI chatbots to continue generating such successful stories. However, we also observed that fans wanted stories unlike the original material (e.g., favoring romance and disliking when characters are hurt), hence AI-powered stories may be less popular if they strictly follow the original material of a show.

Editor's Spotlight/Take 5: How Does ChatGPT Use Source Information Compared With Google? A Text Network Analysis of Online Health Information.

Editor's Spotlight/Take 5: How Does ChatGPT Use Source Information Compared With Google? A Text Network Analysis of Online Health Information.

Do non-motorists understand the traffic safety laws protecting them? Results from a Chinese survey

The government of Nanjing, China, has issued several traffic laws to protect non-motorists (i.e., pedestrians, e-bike riders, cyclists). In reality, however, many non-motorists are unfamiliar with these laws, which impairs the effectiveness of laws. On this background, this study investigates road users’ knowledge on various traffic safety laws protecting non-motorists. An online survey was conducted to collect road users’ basic information and their subjective and objective knowledge of the laws. Then, A latent class analysis model is employed to distinguish between drivers and non-motorists. Furthermore, we explore the gaps between road users’ subjective and objective knowledge, and the knowledge discrepancies between different road users. Subsequently, ordered logit models are developed to identify the influencing factors of road users’ knowledge levels. Finally, text network analysis is used to assess respondent’s comments and suggestions.The results indicate that drivers are more knowledgeable about laws compared with non-motorists. Moreover, some road users have a high level of subjective knowledge but a low level of objective knowledge regarding some laws (e.g., yielding laws). Furthermore, some factors, “group membership” and “reading-news behavior”, significantly affect both subjective and objective knowledge. “Social norm” and “perceived behavioral control” could only affect subjective knowledge, while “perceptions on fine penalty” could influence objective knowledge. Besides, respondents’ positive comments are associated with their perceived benefits from the laws, while the negative comments and suggestions mainly concern non-motorist violators and knowledge discrepancies among different road users. Finally, we provide several targeted suggestions to facilitate the promotion and implementation of traffic safety laws.

How Does ChatGPT Use Source Information Compared With Google? A Text Network Analysis of Online Health Information.

How Does ChatGPT Use Source Information Compared With Google? A Text Network Analysis of Online Health Information.

A Study on the Perception of Destination Image of Myeong-dong, South Korea by Chinese Tourists: Based on Network Text Analysis

A Study on the Perception of Destination Image of Myeong-dong, South Korea by Chinese Tourists: Based on Network Text Analysis

Exploring the Knowledge Structures of Korean and International Nursing Research on Premature Infants Using Text Network Analysis.

This study aimed to identify the knowledge structures of Korean and international nursing studies on premature infants using text network analysis, which represents a text as a network graph that describes how keywords are linked. This network graph refers to a knowledge structure. International and Korean journal databases were searched to extract nursing studies regarding premature infants published in academic journals from 1998 to 2020. Abstracts from the selected studies were analyzed using the following four steps: word extraction and refinement, keyword extraction, co-occurrence matrix generation, and text network visualization. The results demonstrated that 182 Korean and 2502 international studies were published. The common keywords of Korean and international studies were "kangaroo mother care," "stress," and "child." The keywords of the international studies had more branches linking to other keywords than those of the Korean studies. Thus, the knowledge structure of international studies included diverse concepts. These findings will serve as important guidance for future research worldwide. Furthermore, studies to develop a more comprehensive knowledge structure of international research on premature infants are needed. The knowledge structure of Korean studies mainly included concepts related to mothers. Korean studies regarding hospitalized premature infants and communication with parents need to be conducted.

The Impact of Network Structure on Knowledge Adoption: A Network Text Analysis on Knowledge-Sharing Platforms

In recent years, the public has increasingly relied on online knowledge-sharing platforms as credible and valuable sources to acquire information and knowledge. This work employed the network text analysis approach to explore the determinants of knowledge adoption from a structural perspective. First, we proposed a quantitative method to measure knowledge adoption behavior on a collective level based on the knowledge network. Then, we discussed knowledge concept adoption and knowledge relationship adoption. To build knowledge networks and analyze their structural features, this work collected 74 761 knowledge concepts from the Zhihu platform and 62 368 knowledge concepts in the Stack Overflow platform. The regression analysis results showed that the structural characteristics of knowledge concepts and knowledge relationships substantially affect their adoption. This study advances our understanding of knowledge adoption in an online knowledge-sharing platform and provides a structural analysis approach to large-scale online content data.

Improving sentiment analysis using text network features within different machine learning algorithms

Sentiment analysis poses a significant challenge due to the inherent subjectivity of natural language and the prevalence of unstandardized dialects in social networks. Regrettably, existing literature lacks a dedicated focus on network representation learning for sentiment classification. This paper addresses this gap by investigating ten machine learning algorithms, including support vector machine (SVM), random forest (RF), logistic regression (LR), and Naive Bayes (NB). Our approach integrates text network analysis and sentiment analysis to propose a comprehensive solution. We begin by applying text preprocessing techniques and converting a text corpus into a text network using word co-occurrence. Subsequently, we employ network analysis techniques to extract features based on network topology and node attributes. These network-derived features serve as inputs for sentiment prediction on Yelp reviews. Through the incorporation of diverse text network features and various machine learning algorithms, we achieve significant enhancements in sentiment classification performance. Our evaluation demonstrates an improved area under curve (AUC) of 83% on the Yelp reviews corpus, underscoring the efficacy of integrating network features to enhance sentiment classifiers. This research underscores the critical role of network representation and its potential impact on sentiment analysis, highlighting the prospect of harnessing network features for sentiment classification tasks.