This study conducts a comparative analysis of various machine learning and deep learning models for predicting order quantities in supply chain tiers. The models employed include XGBoost, Random Forest, CNN-BiLSTM, Linear Regression, Support Vector Regression (SVR), K-Nearest Neighbors (KNN), Multi-Layer Perceptron (MLP), Recurrent Neural Network (RNN), Bidirectional LSTM (BiLSTM), Bidirectional GRU (BiGRU), Conv1D-BiLSTM, Attention-LSTM, Transformer, and LSTM-CNN hybrid models. Experimental results show that the XGBoost, Random Forest, CNN-BiLSTM, and MLP models exhibit superior predictive performance. In particular, the XGBoost model demonstrates the best results across all performance metrics, attributed to its effective learning of complex data patterns and variable interactions. Although the KNN model also shows perfect predictions with zero error values, this indicates a need for further review of data processing procedures or model validation methods. Conversely, the BiLSTM, BiGRU, and Transformer models exhibit relatively lower performance. Models with moderate performance include Linear Regression, RNN, Conv1D-BiLSTM, Attention-LSTM, and the LSTM-CNN hybrid model, all displaying relatively higher errors and lower coefficients of determination (R²). As a result, tree-based models (XGBoost, Random Forest) and certain deep learning models like CNN-BiLSTM are found to be effective for predicting order quantities in supply chain tiers. In contrast, RNN-based models (BiLSTM, BiGRU) and the Transformer show relatively lower predictive power. Based on these results, we suggest that tree-based models and CNN-based deep learning models should be prioritized when selecting predictive models in practical applications.

Despite the surge of publication of chatbots in the recent years in the field of education, we have little to know how this area has been researched so far, and the metrics of this type of research is still not known. To address such gap, this article offers a descriptive bibliometric study of chatbot research in education, aiming at presenting bibliometric analysis on articles on chatbots in education that were published in journals indexed in the Web of Science (WOS) database specifically Social Science Citation Index (SSCI) and Science Citation Index Expanded (SCIE) between 2016 and 2023. Descriptive bibliometric analysis was used to examine the data gathered from the chosen publications. including the annual number of articles and citations, the most productive author, countries with the highest publication output, productive affiliations, funding organizations, and publication sources. The bulk of the articles on chatbots in education, according to our dataset, were published between 2016 and 2023. The United States of America tops the list of countries regarding research productivity. The United Kingdom and China were ranked as most second and third productive countries, in terms of publication outputs. “Luke Kutszik Fryer emerged as the most productive author in this research domain in terms of the number of publications.” The University of Hong Kong had the highest number of publications among affiliations, indicating their significant contribution to the field. Additionally, the journal “Computers in Human Behavior” stood out with the highest number of publications per year, highlighting its relevance in publishing research on chatbots in education. This research offers valuable insights and a roadmap for prospective researchers, pinpointing critical areas where success can be attained in the study of chatbots in education.

The quest for quality postgraduate research productivity through education is on the increase. However, in the context of the African society, governance structures and policies seem to be impacting on the quality level of the provided education. Hence, this conceptual study explored the roles of governance structures and policies in enhancing and ensuring quality postgraduate education programmers in African institutions of higher learning. To this end, various relevant literature was reviewed. The findings showed amongst others that governance structures and policies affect the quality of education provided. Meanwhile, other factors such as curriculum, foreign influence, lack of resources, training, amongst others contribute to the quality of education provided. The study concludes that there is need for the current structures of governance and the designed and implemented policies for postgraduate education to be reviewed and adjusted towards ensuring the desired transformation.

 Distance education (DE) has recently become a noteworthy study topic in the public education system. From the Web of Science database, 5719 articles discussing DE and published in the period of 2011–2023 were acquired. By analyzing the overall characteristics, co-citation, and keyword co-occurrence of the selected articles, which utilized Cite Space software, the history of DE could be systematically grasped, thereby reasonably predict the emphases of future development. We found that the number of papers relevant to DE had been rapidly growing since 2018. USA, China, and Turkey are the top three countries where most authors or teams were located. The map of keyword cooccurrence showed that the previous DE research mainly focused on telelearning, adult learning, and distributed learning environment. The recent burst words emerging are used to determine that distance education will continue to be studied in the field with high explosive keywords such as visual tracking, technology acceptance model, and user interface. This will provide suggestions and directions for the development of distance education.

Given the issues of urban-rural educational inequality and difficulties for children from poor families to succeed, this study explores the impact mechanism of internet usage on rural educational investment in China within the context of the digital divide. Using data from the 2019 China Household Finance Survey (CHFS), this study analyzed the educational investment decisions of 2064 rural households. Results indicate that in the Eastern region, a high level of educational investment is primarily influenced by the per capita income of the family, with social capital and internet usage also playing supportive roles. In the Northeastern region, the key factor is the diversity of internet usage, specifically using both a smartphone and a computer. In the Central region, factors such as the diversity of internet usage, subjective risk attitudes, the appropriate age of the household head, and per capita income of the family contribute to higher levels of educational investment. In the Western region, the dominant factors are the diversity of internet usage, subjective usage and per capita income of the family. These factors enhance expected returns on the high level of educational investment and boost farmers’ confidence. High internet usage rates significantly promote diverse and stable educational investment decisions, providing evidence for policymakers to bridge the urban-rural education gap.

STEAM (science, technology, engineering, arts, and mathematics) education has recently been encouraged and attracted much national attention. This qualitative study aimed to conduct a thematic analysis of college student STEAM open responses to provide an examination of college students’ perceptions of their STEAM experiences into the STEAM field. Based on transformative learning theory, a thematic analysis of 756 written responses to seven prompts by 108 college student participants revealed three primary themes: (1) exciting and challenging difficulties, and transdisciplinary learning in STEAM; (2) STEAM learning of gradual process, problem-oriented instruction, and creative problem solving; and (3) metacognition development in STEAM. The findings revealed that undergraduates’ STEAM perceptions provide strong support for STEAM implementation to enhance teaching effectiveness in higher education.