The urbanization process has led to a significant increase in energy consumption and carbon emissions, which can be mitigated through scientific urban planning and management. This research proposes a bottom-up urban carbon emission mitigation strategy based on deep reinforcement learning (DRL). Using Ningbo City as a case study, multi-source urban data, including points of interest (POI) data and urban transportation system data, are utilized, along with varying carbon emission coefficients for different travel modes, to construct a comprehensive carbon emission environment for urban areas. The proposed DRL model adopts an Actor-Critic framework, which iteratively optimizes the land use configuration and building type proportions within the urban matrix to achieve the goal of mitigating travel carbon emissions. Experimental results demonstrate that this approach exhibits significant carbon reduction effects in urban scenario. By adjusting the discount rate of the reward function, various optimization strategies can be obtained, such as short-term and long-term strategies, achieving reductions of 0.47% and 0.61%, respectively, which are notably higher than the 0.39% reduction expected if travel emissions were uniformly distributed across the matrix. The findings highlight the potential of DRL-based approaches in urban planning to achieve adaptive and data-driven strategies for carbon emission mitigation.

Urban land-use optimization plays a vital role in mitigating the escalating carbon emissions of rapidly growing cities. This study employs advanced computational intelligence to address urban carbon reduction through optimized spatial configurations. A Deep Reinforcement Learning (DRL) framework is proposed that integrates Points of Interest (POI), Areas of Interest (AOI), and Transportation System Data (TSD) to generate fine-grained carbon emission maps guiding land-use adjustments. In the case study of Hangzhou, China, results show that a carefully designed reward function enables the DRL agent to selectively optimize land-use structures, prioritizing the centralization of residential, dining, and commercial areas to form high-density, mixed-use urban clusters. This spatial reorganization leads to notable reductions in carbon emissions and improvements in resource-use efficiency. The proposed DRL-based framework provides a scientific basis for policy development toward sustainable land-use and urban density optimization. By merging advanced AI techniques with urban planning, this research contributes to the creation of low-carbon, resilient, and environmentally sustainable cities capable of addressing global climate challenges. The optimized DRL agent achieved carbon emission reductions of up to 15% compared to baseline configurations in the Hangzhou case study. Spatial concentration analysis revealed a 23% increase in residential area clustering and 31% increase in commercial zone centralization over 400 training episodes. The PPO-based model demonstrated superior performance compared to genetic algorithm and linear regression baselines, with lower policy loss (converging to <0.01) and critic loss (converging to <0.005) after early stopping at 400 episodes. However, this study is limited by its deterministic environment model, geographic specificity to Hangzhou, and exclusive focus on carbon reduction without incorporating socioeconomic constraints.

Accelerating urban photovoltaic (PV) deployment is central to decarbonisation, yet allocating large-scale projects across urban land involves competing governance and spatial-planning demands. Existing approaches rely either on rule-based potential maps or on building-scale models tied to roof morphology. We first compare replicated theoretical potential surfaces with realised PV deployment across six Australian urban regions and show that observed sites cluster mainly within the top 50%, rather than the theoretical top 10%, of urban land by PV potential. We then introduce SPADE (Spatial Photovoltaic Allocation via Deep Learning and Explainability), a morphology-free framework that learns macro-scale siting logic directly from observed deployment patterns and multi-source spatial context. SPADE couples convolutional encoding with Transformer attention to represent both local structure and wider urban dependencies across diverse rasterised inputs. Trained on 939 PV installations from five Australian cities and transferred without retraining to the Australian Capital Territory (ACT), SPADE reproduces 64.6% of existing sites within 1 km, including 6.52 MW of installed capacity, while delineating a predicted suitable area covering only 9.69% of urban land. Interpretation analyses further show that 16.13% of ACT sites fall within larger high-suitability clusters with scope for intensification, whereas 25.8% deviate from the dominant learned pattern, suggesting additional institutional or strategic drivers. By jointly diagnosing mismatch between theoretical suitability and realised deployment and recovering a transferable, inspectable siting logic without building morphology, SPADE offers an empirical basis for more transparent and reviewable urban renewable-energy planning.

Digital media platforms have become central to how people perceive nations, cultures, and global issues. Yet existing research rarely captures how content visibility and emotional tone jointly shape audience responses in UGC platform. Traditional survey or content-analytic methods identify correlations but overlook the nonlinear interactions now traceable through large-scale platform data. Using China-related videos on YouTube as a case study, this research propose a computational framework that links media content, exposure dynamics, and audience sentiments within a unified modeling pipeline. It integrates RoBERTa- and BART-based sentiment quantification with exposure indicators and predict average comment sentiment (ACS) using a multilayer deep neural network (DNN) enhanced with explainable AI (XAI). Results show that Exposure metrics, including view count (23.78%), comment count (20.46%), and subscriber count (18.91%), emerge as the dominant determinants of ACS. SHAP dependence plots indicate an overall positive association between Content Valence and ACS. Exposure metrics display different patterns: views are positively associated with ACS, whereas comment counts are negatively associated. At low exposure levels, ACS and content valence are aligned in polarity; at high exposure levels, the effect weakens or reverses, and negative-valence videos exhibit a U-shaped pattern. An additional finding is that, during medium-to-high exposure periods, videos from high-subscriber accounts maintain persistently negative ACS, whereas those from mid-subscriber accounts show polarity reversals. This research offers a new methodological perspective for computational communication research, extends the application of Second-Level Agenda-Setting to UGC contexts, and provides practical insights for communication strategy.