Search

Pre-evaluating scientific paper quality aids in alleviating peer review pressure and fostering scientific advancement. Although prior studies have identified numerous quality-related features, their effectiveness and representativeness of paper content remain to be comprehensively investigated. Addressing this issue, we propose a content-based interpretable method for pre-evaluating the quality of scientific papers. Firstly, we define quality attributes of computer science (CS) papers as integrity, clarity, novelty, and significance, based on peer review criteria from 11 top-tier CS conferences. We formulate the problem as two classification tasks: Accepted/Disputed/Rejected (ADR) and Accepted/Rejected (AR). Subsequently, we construct fine-grained features from metadata and knowledge entity networks, including text structure, readability, references, citations, semantic novelty, and network structure. We empirically evaluate our method using the ICLR paper dataset, achieving optimal performance with the Random Forest model, yielding F1 scores of 0.715 and 0.762 for the two tasks, respectively. Through feature analysis and case studies employing SHAP interpretable methods, we demonstrate that the proposed features enhance the performance of machine learning models in scientific paper quality evaluation, offering interpretable evidence for model decisions.

We investigate the sub-networks involved in education policy in Israel in recent years, using Mixed Methods Social Networks Analysis - drawn from combined analysis of qualitative and quantitative data. Our objective was to comprehensively explore the Israeli education policy network to deliver an understanding of its structure, actors, and relationships. Our research offers a descriptive, analytical, and interpretive account of the contemporary Israeli education policy sector, including the actors involved, central policy sub-networks and organizations, and their relationships. We analyze dependence relations within the sub-networks using Resource Dependency Theory and Policy Networks Typology. Our findings contribute to understanding the dynamics of Israeli education policy networks and their perceived influence on policy-making and enactment processes. To the international field of educational policy-related research, we introduce a novel category of policy network actors, previously unaddressed as a distinct type. We have termed this category 'Ethos networks' to characterize this group of actors and elaborate on its significance within the field. Additionally, we underline the importance of considering external societal and political factors in education policy-making.

Pre-evaluating scientific paper quality aids in alleviating peer review pressure and fostering scientific advancement. Although prior studies have identified numerous quality-related features, their effectiveness and representativeness of paper content remain to be comprehensively investigated. Addressing this issue, we propose a content-based interpretable method for pre-evaluating the quality of scientific papers. Firstly, we define quality attributes of computer science (CS) papers as integrity , clarity , novelty , and significance , based on peer review criteria from 11 top-tier CS conferences. We formulate the problem as two classification tasks: Accepted/Disputed/Rejected (ADR) and Accepted/Rejected (AR). Subsequently, we construct fine-grained features from metadata and knowledge entity networks, including text structure, readability, references, citations, semantic novelty, and network structure. We empirically evaluate our method using the ICLR paper dataset, achieving optimal performance with the Random Forest model, yielding F1 scores of 0.715 and 0.762 for the two tasks, respectively. Through feature analysis and case studies employing SHAP interpretable methods, we demonstrate that the proposed features enhance the performance of machine learning models in scientific paper quality evaluation, offering interpretable evidence for model decisions. • Define four criteria for quality evaluation of scientific papers: integrity, clarity, novelty, and significance. • Propose a framework for quality evaluation of scientific papers based on coarse features and knowledge entity network. • An effective algorithm for measuring the novelty and significance of scientific papers based on knowledge entity networks. • Create and release a rigorous dataset, which could serve as the gold standard for quality evaluation of scientific papers. • Conduct extensive experiments to validate the effectiveness of the proposed framework. [ABSTRACT FROM AUTHOR]

Background: Learner profiles detected from digital trace data are typically triangulated with survey data to explain those profiles based on learners' internal conditions (e.g., motivation). However, survey data are often analysed with limited consideration of the interconnected nature of learners' internal conditions. Objectives: Aiming to enable a thorough understanding of trace-based learner profiles, this paper presents and evaluates a comprehensive approach to analysis of learners' self-reports, which extends conventional statistical methods with psychological networks analysis. Methods: The study context is a massive open online course (MOOC) aimed at promoting physical activity (PA) for health. Learners' (N = 497) perceptions related to PA, as well as their self-efficacy and intentions to increase the level of PA were collected before and after the MOOC, while their interactions with the course were logged as digital traces. Learner profiles derived from trace data were further examined and interpreted through a combined use of conventional statistical methods and psychological networks analysis. Results and Conclusions: The inclusion of psychological networks in the analysis of learners' self-reports collected before the start of the MOOC offers better understanding of trace-based learner profiles, compared to the conventional statistical analysis only. Likewise, the combined use of conventional statistical methods and psychological networks in the analysis of learners' self-reports before and after the MOOC provided more comprehensive insights about changes in the constructs measured in each learner profile. Major Takeaways: The combined use of conventional statistical methods and psychological networks presented in this paper sets a path for a comprehensive analysis of survey data. The insights it offers complement the information about learner profiles derived from trace data, thus allowing for a more thorough understanding of learners' course engagement than any individual method or data source would allow.

In 2014, the OECD-PISA's Governing Board approved the addition of a set of global competence measures to its Programme of Student Assessment. In our paper, we explore whether and how there are discursive shifts between the two framing papers (2016/2018) and what the outcomes are for policy-shaping. To this end, we employ Network Text Analysis to map shifting semantic configurations. We show that (i) the concept of 'global competence' is radically redefined through the simplification and polarisation of the semantic universe surrounding it, (ii) that the concept 'global' becomes a shifting signifier which enables the establishment of an equivalence between the two studied documents, and that (iii) in this process, concepts such as 'culture' are now erased in the 2018 text. Our findings show the dramatic change of approach between the two documents reinforces a narrative that is familiar to the OECD around knowledge economy proxies.

Purpose: Design-based research (DBR) involves multiple iterations, and innovations are needed in analytical methods for understanding how learners experience a learning experience in ways that both embrace the complexity of learning and allow for data-driven changes to the design of the learning experience between iterations. The purpose of this paper is to propose a method of crafting design moves in DBR using network analysis. Design/methodology/approach: This paper introduces learning experience network analysis (LENA) to allow researchers to investigate the multiple interdependencies between aspects of learner experiences, and to craft design moves that leverage the relationships between struggles, what worked and experiences aligned with principles from theory. Findings: The use of network analysis is a promising method of crafting data-driven design changes between iterations in DBR. The LENA process developed by the authors may serve as inspiration for other researchers to develop even more powerful methodological innovations. Research limitations/implications: LENA may provide design-based researchers with a new approach to analyzing learner experiences and crafting data-driven design moves in a way that honors the complexity of learning. Practical implications: LENA may provide novice design-based researchers with a structured and easy-to-use method of crafting design moves informed by patterns emergent in the data. Originality/value: To the best of the authors' knowledge, this paper is the first to propose a method for using network analysis of qualitative learning experience data for DBR.

Teach for Australia was announced by the Australian Government in 2008, at a corporate dinner sponsored by Swiss multinational investment bank UBS, hosting New York Schools Chancellor Joel Klein. Conceptualising Teach for Australia as a polycentric policy network anchored in venture philanthropy, this paper examines how networks mobilise major government reform, drawing upon both market and state. As an organisation founded, staffed and registered by a major global consultancy firm (Boston Consultancy Group), the paper critiques and questions representation, divisibility and immutability. By drawing upon network ethnography, following people and policy, the paper traces the beginnings of Teach for Australia through inaugural founders, focusing upon co-affiliations. Mapping a heterarchical and polycentric network of non-state and state actors, it highlights indivisible flows of people, money and policy. Thus, this paper seeks to indicate how policy networks embedded in venture philanthropy mobilise policy, whilst being reliant upon formal bureaucracies. This is governance with government. Teach for Australia highlights the nexus between private and public, and fuses together critical connections between market and state, bolstering a major venture philanthropic network that chiefly impacts public schools.

The paper reports about the outcome of a systematic review of research on family resemblance approach (FRA) to nature of science in (NOS) science education. FRA is a relatively recent perspective on NOS being a system of cognitive-epistemic and social-institutional aspects of science. FRA thus consists of a set of categories such as aims and values, practices, knowledge and social organizations in relation to NOS. Since the introduction of the FRA, there has been increasing interest in investigations about how FRA can be of use in science education both empirically and practically. A journal content analysis was conducted in order to investigate which FRA categories are covered in journal articles and to identify the characteristics of the studies that have used FRA. These characteristics included the target level of education and focus on pre- or in-service teachers. Furthermore, epistemic network analysis of theoretical and empirical papers was conducted to determine the extent to which the studies incorporated various key themes about FRA, such as its transferability to other domains and differentiation of the social-institutional system categories. The findings illustrate an increasing number of empirical studies using FRA in recent years and broad coverage in science education. Although the social-institutional system categories included intraconnections, these were not as strong as those intraconnections among categories within the cognitive-epistemic system. Future research directions for the use of FRA in K-12 science education are discussed.

Since social media is increasingly pervasive in modern society, this bibliometric study aims to investigate its educational applications. Using the Scopus database, the bibliometric method analyses publications published between 2010 and 2022. The research indicates that student participation and ease of access are the two main benefits of using social media in the classroom. However, it also spreads misinformation and poses privacy and security risks. Articles that discussed how social media could be used in the classroom were found and organised using a bibliometric analysis based on their subject matter, year of publication, and authors. The research shows that between 2001 and 2020, there was a rise in the number of papers discussing the use of social media in the classroom. In addition, the top five countries in terms of annual publication output include the United States, the United Kingdom, Australia/India, and Canada. To further explore the connections between relevant articles, a co-citation network analysis was performed. Therefore, there must be strict rules and policies for using social media in education to address privacy and security concerns and the spread of false information.

Massive open online courses (MOOCs) have also received interest from researchers worldwide; however, there was no comprehensive review of the MOOC research. This paper aims to identify the MOOCs research scientific landscape as the trend from publications worldwide. In assessing research trends, the bibliometric network analysis using distance-based network mapping in VOSviewer was applied in this review. The 3,211 eligible articles published between 2011 and 2020 confirmed three main research clusters: learning system, human characteristics and higher education clusters. The results also showed that terms, such as 'learning systems', 'gender differences' and 'flipped classroom' emerged as ongoing research trends. In addition, these findings indicated that the overall productivity rates in the Middle East and Gulf regions were low. Besides, the authorship mapping indicated an absence of the small-world properties. A discussion of the findings and directions for further research are also provided. Based on the network analysis method, this paper presents the researchers' alternative method to review literature using an approach that possibly includes mostly published articles related to MOOCs.

Aim: A well-known method for quantitatively evaluating scholarly work is bibliometric analysis. Best-cited papers raise awareness of the influential publications and patterns in the literature on a specific subject. The aim was to conduct bibliometric analysis to determine most cited articles on vesiculobullous oral lesions. This is the first study on citation analysis with respect to vesiculobullous oral lesions. Materials and Methods: A retrospective data search was explored on December 2022 using the Scopus database. The articles were evaluated, and fundamental data for bibliometric analysis was reviewed. Standard details about the author, linked organizations, publishing year, and place of origin were noted. Statistical analysis was performed using Chi-square analysis. VOSviewer software was used to determine the bibliometric network analysis for co-occurrence among coauthors and commonly used keywords. Results: A total of 344 articles published from 1971 to 2022 were included in the study. A total of 6680 citations and 19.41 citations per article were observed. The journal Archives of Dermatology received the most citation. There was a significant association between the number of citations and the journal type (open access vs. non-open access) (P < 0.05). Four to five highly related clusters with the help of VOSviewer software were found during co-occurrence network analysis. Conclusions: The top 10 articles on vesiculobullous oral lesions that received the most citations were listed in detail in the present study. This will be a valuable resource for academics, clinicians, and researchers in the fields of dermatology, general pathology, oral pathology, and oral medicine. [ABSTRACT FROM AUTHOR]

Humans have always been lured by the idea that they can use data to understand a phenomenon and make predictions about it. Learning analytics, in this sense, promise to understand and optimize learning and the environments in which it occurs by collecting data from learners and learning contexts. In this regard, this study systematically examines research on learning analytics through bibliometric, data mining, and analytics approaches. This paper argues that research interest in learning analytics is increasing steadily; some countries, higher education institutions, and researchers have a specific research agenda that indicates their intention to specialize in that field. It is also noted that there is a need for more interdisciplinary studies on learning analytics and a further need to merge technological capabilities with pedagogy. Based on the findings of text-mining and social network analysis, the following themes were identified: (1) learning analytics to improve teaching and personalize learning, (2) hegemony of data-driven teaching and learning practices, (3) multimodal learning analytics as the next generation practice, (4) learning design for learning analytics, (5) formative assessment through learning analytics, (6) learning analytics for social online learning spaces, and (7) privacy and ethical concerns to overcome. This paper suggests focusing on issues such as ethics and privacy and warns researchers to pay attention to the risks of both an educational panoptic society and quantified decision-making processes. Furthermore, rather than relying on algorithms, it is suggested to incorporate social values and center the learners in the learning analytics processes. Finally, this paper asks the following: "If we quantify the learning processes, how can we benefit from the carpe diem of educational processes and then seize the beauty of teaching and learning?"

The interdisciplinary field of human movement sciences (HMS) has gained massive interest among educational institutions around the world, not only in terms of academic programs but also in research. With this emergence, the researchers aimed to describe the productivity of HMS research in the Philippines. The descriptive bibliometric analysis phase of this paper considered papers published and indexed in Google Scholar from January 2010 to June 2021 and was analyzed after data cleaning and preprocessing. Results revealed that a total of 274 research publications were recorded between the years 2010 and 2021 with an average annual publication rate of 28.6% as far as the dataset was concerned. Also, public higher education institutions (HEIs) emerged to be the most productive generators of research outputs, specifically topped by the University of the Philippines (UP). Moreover, UP-based authors dominated the rankings of the most productive HMS researchers, with "Jeffrey Pagaduan" in the top rank. Results further indicated that 75% of the authors collaborated with fellow researchers within or outside their institution. Meanwhile, the term "physical education" was recorded to be the most frequently appearing word in most of the publications. Through the aid of the data mining approach social network analysis (SNA), five (5) researchers from public HEIs and two from private HEIs were identified to have the largest networks as discussed in the paper. The exploration of research in terms of productivity and the embedded social network of researchers may serve as a springboard in defining the future development of the field in the Philippines.

Understanding complex systems is important in science and has become a prominent aspect of STEM education in recent years. System thinking plays a critical role, focusing on system attributes such as interactions between components, emerging phenomena and hierarchical organisation. In this study, we provide a network analysis of bibliometric information and a research synthesis of empirical studies on system thinking and complexity in STEM education from the past two decades, aiming to identify trends in the field. We reviewed 255 empirical published studies identified from Web of Science database and found a sharp increase in the number of studies on system thinking after 2016. In the research synthesis, we categorised these studies under five categories: (i) study population, (ii) disciplinary field, (iii) empirical study type, (iv) system attributes and (v) cognitive aspects. In the results, we found that most studies focused on higher education (86 papers) and biology (67 papers), while only few of the studies focused on pre- and in-service teachers (14 and 23 papers, respectively) as well as on elementary students (22 papers). Complexity and interactions emerged as the most mentioned system attributes (89 and 55 papers, respectively), while thinking and understanding were the most mentioned cognitive aspects (112 and 73 papers, respectively). In addition, an increasing focus on digitalisation (68 papers) and modelling (61 papers) was identified. We discuss these emerging trends and identify research gaps and directions for future studies.

This paper presents a cartography of the digital literacy academic field. Such cartography is comprised of two sections: a categorization of the field through literature review and analysis, and an exploration of its main issues through thematic and network analysis. On the one hand, five conceptual categories of digital literacies are found: functional, sociocultural, critical, transformative, and sociomaterial. On the other, main issues are described with 21 recurring themes of digital literacy and a few networks depicting its most salient matters of concern, concluding with an interpretation of these in the composition of 8 encompassing issue spaces: digital literacies conceptions and practices, digital literacy in education, access and digital divide, digital texts and literacy, websites and social networks, digital technologies at the workplace and healthcare, digital technologies users and uses, and information issues. Finally, a few paragraphs are dedicated to the limitations of categorizing and issue mapping.

I trace the bibliometric evolution of "New Public Management Is Dead" by Dunleavy et al. to investigate how the seminal paper influenced the administrative reform debate. They suggested Digital-Era Governance as the main post-NPM idea. My bibliometric analysis discovers public value, administrative reform trajectories, and digital government as influential themes. Unlike Dunleavy et al., the literature found the managerial reform wave is not linear, reform ideas are supplementary, and NPM remains a major toolkit. Future research should focus on reintegration and need-based holism, linking digital government to administrative reform, and the negative impact of digital government on democracy. [ABSTRACT FROM AUTHOR]