The United Nations Sustainable Development Goals (SDGs), established in 2015, are binding targets for monitoring the sustainable status of all countries in the world. Developing localized SDGs indicator systems, assessing SDGs progress comprehensively, and creating policy tools have gradually become the hotspots of scientific research and practical application of the SDGs. To systematically sort out the research status of the SDGs progress and identify problematic gaps, this paper uses bibliometric methods to analyse the scientific knowledge mapping of SDGs research, sort out the general characteristics, scientific cooperation, cutting-edge hotspots and future research trends of SDGs research. Besides, this paper promotes the integration of academic research into concrete practice by linking it to mainstream SDGs progress reports. The results show that: ① The amount of SDGs research literature in 2015–2022 is on a growing trend, and the existing research is characterised by multidisciplinary crossover. Research institutions such as Utrecht University, the Chinese Academy of Sciences, and the University of the Chinese Academy of Sciences have contributed the most to SDGs research. ② The research hotspots include the SDGs practice of different responsible subjects, the comprehensive assessment of the progress of SDGs, the scientific research of SDGs indicators, and the research on the interaction between the goals of SDGs. ③ Environmental issues, indicators, energy transition, education and agriculture are the hot directions of SDGs research. Energy saving and carbon reduction, sustainable practices of supply chain management, and promoting the progress of environmental dimension goals are the focus of subsequent research. According to the results of the bibliometric analysis, future SDGs research should pay attention to strengthening multi-party cooperation, using innovative technologies to support the assessment of SDGs progress, and formulating sustainable development strategies.

CO₂ emission reduction has become a consensus all around the world. This paper investigates the spatiotemporal characteristics of industrial land carbon emission intensity (ILCEI) in China by spatial autocorrelation analysis, and applies the spatial Durbin model to reveal the influence of driving factors on ILCEI. The results indicate the following: (1) national ILCEI first shows a downward and then an upward trend during the period and presents a low pattern in both Eastern and Northeastern regions and a high pattern in the Northwestern region. (2) From a global perspective, ILCEI shows significant spatial agglomeration characteristics; from a local perspective, ILCEI is dominated by H-H and L-L agglomeration types, showing that spatial heterogeneity and spatial dependence are apparent in ILCEI. (3) ILCEI is significantly negatively affected by both R & D personnel and foreign-trade dependence, while urban population density notably has positive impacts on ILCEI. This paper is a beneficial policy practice for harmonizing the contradiction between industrial land expansion and carbon discharge.

In the new stage of China’s economic transformation from high-speed growth to high-quality development, scientific and technological factors have become the new driving force for the high-quality development of China’s economy. Therefore, how the flow of scientific and technological factors affects the high-quality development of regional economy is worthy of in-depth discussion. Based on the panel data of 30 provinces (municipalities and districts) in China from 2006 to 2019, this paper calculates the high-quality economic development index by constructing a comprehensive index system, and empirically examines the direct effect, spatial effect and threshold effect of the flow of scientific and technological factors (including scientific and technological talent factors and scientific and technological capital factors) on high-quality regional economic development. The results show the following: (1) The flow of science and technology talents and science and technology capital is the main driving factor for the high-quality development of regional economy, and the flow of science and technology talents and science and technology capital can play a complementary role. (2) High-quality regional economic development in China presents a certain clustering phenomenon in geographical space. The flow of scientific and technological talent elements and scientific and technological capital elements can stimulate the improvement of high-quality economic development in surrounding areas through the spatial spillover effect of high-quality economic development. (3) Only when the flow of science and technology factors is matched with a higher degree of financial structure and industrial structure can the flow of science and technology factors more effectively promote the high-quality development of regional economy. (4) The higher the level of industrialization and economic development, the more obvious the impact of the flow of scientific and technological talent factors and scientific and technological capital factors on the high-quality economic development. Meanwhile, the lower the degree of marketization, the more obvious the impact of the flow of scientific and technological capital factors on the high-quality economic development.

In the context of increasingly complex global economic and political environments, accurately predicting macroeconomic indicators, especially the Consumer Price Index (CPI), is crucial for economic stability and policy formulation. This study focuses on using neural network models, particularly those enhanced with multi-layer attention mechanisms such as Long Short-Term Memory (LSTM) networks, to analyze and predict key macroeconomic indicators like CPI. By integrating natural language processing techniques, a comprehensive predictive dataset was constructed for training and testing the model. The study reveals that the multi-layer attention mechanism significantly improves the model's capability in capturing critical economic events and cyclical changes, particularly in analyzing various factors affecting CPI. Additionally, the enhanced LSTM model was compared and validated against other machine learning models, demonstrating its superiority in CPI forecasting. The application of text mining techniques further enhances the model's ability to grasp market dynamics, making CPI predictions more timely and accurate. The methods and findings of this study not only provide a new perspective for predicting macroeconomic indicators like CPI but also offer valuable decision-support tools for policymakers and market analysts.