More and more tourists tend to seek cultural enjoyment while taking trips based on their satisfied socio-economic needs. The spatiality of cultural tourism determines the feasibility and convenience of the access to cultural tourism resources. Given that spatial research on cultural tourism rarely involves the underdeveloped regions in China, especially via Baidu Index data, this study aims to explore the dynamic spatial pattern of cultural tourism Internet attention based on Baidu Index through social network analysis (i.e., network density, centrality analysis, core–periphery model, etc.) in the Western Triangle Economic Zone (WTEZ), China. Research findings are as follows. (1) Temporally, the centrality is increasingly enhanced in terms of the core cities with high cultural tourism Internet attention, as well as their radiation effect. (2) Spatially, the distribution pattern of cultural tourism Internet attention is shifting from polarization to equilibrium. However, the overall network density still presents a mode of loose connection and fails to extend from the core triangle structure in the planar dimension. (3) The cultural tourism Internet attention in WTEZ presents an obvious core–periphery pattern. Xi’an–Chengdu–Chongqing is in a dominant position, and the core area spreads from south to north. (4) The spatial-temporal pattern of cultural tourism Internet attention in WTEZ is influenced by transportation, reception facilities, consumption ability, and political environment, and the main driving factors are the number of hotels, per capita income, total highway mileage, and cultural tourism policy indicators. This study is conducive to sustainable spatial planning of cultural tourism through smart data governance.

Despite China's building into a leading sporting nation and sport-tourism integration high-quality development strategy in the 20th National Congress of the Communist Party of China, existing tourism studies seldom concern sport-tourism integration, especially their spatial hot spots and evolutional trend based on geospatial big data. This study aims to probe into the spatiality and the underlying mechanism of sport tourism through internet attention data in 2015, 2018, and 2021 by social network analysis, with a specific focus on aero-sports tourism. Shaanxi Province is chosen as the study site given its advantages of rich aero-sports tourism resources and various aero-sports modes (e.g., sky diving, paragliding, etc.). The results are concluded: (1) At the provincial scale, the aero-sports tourism internet attention shows a pattern of “strong in the middle and weak in the north and south”. (2) At the regional scale, the sub-group clusters within the three specific regions (Shanbei, Guanzhong, and Shannan) of Shaanxi Province turn to be inter-regional clusters. Guanzhong region, especially with Xi'an as the core, is dominant in connecting its peripheral area. Since 2016, the radiation effects of the Guanzhong Region have shown a homogeneous trend of yearly growth and effect strengthening, yet become loosely connected with a heterogeneous trend in 2021 due to the COVID-19 epidemic. (3) At the city scale, the core area of aero-sports tourism internet attention expanded from Xi'an and Xianyang to Yulin and Baoji from 2015 to 2021, resulting from urban economic strength and aviation flight camp club development. (4) The number of general aviation manufacturers, tourist attendance, and tourism revenue significantly affect aero-sports tourism internet attention.

Data-driven deep neural networks have demonstrated their superiority in high-resolution remote-sensing image (HRSI) classification based on superpixel-based objects. Currently, most HRSI classification methods that combine deep learning and superpixel object segmentation use multiple scales of stacking to satisfy the contextual semantic-information extraction of one analyzed object. However, this approach does not consider the long-distance dependencies between objects, which not only weakens the representation of feature information but also increases computational redundancy. To solve this problem, a superpixel-based long-range dependent network is proposed for HRSI classification. First, a superpixel segmentation algorithm is used to segment HRSI into homogeneous analysis objects as input. Secondly, a multi-channel deep convolutional neural network is proposed for the feature mapping of the analysis objects. Finally, we design a long-range dependent framework based on a long short-term memory (LSTM) network for obtaining contextual relationships and outputting classes of analysis objects. Additionally, we define the semantic range and investigate how it affects classification accuracy. A test is conducted by using two HRSI with overall accuracy (0.79, 0.76) and kappa coefficients ($\kappa$) (0.92, 0.89). Both qualitative and quantitative comparisons are adopted to test the proposed method’s efficacy. Findings concluded that the proposed method is competitive and consistently superior to the benchmark comparison method.