Search

This study investigates the dynamic relationship between cross-border e-commerce and logistics co-development strategies within the context of small and medium-sized enterprises (SMEs) in China. The primary objective is to formulate an empirical model capable of estimating the utility and flexibility of cross-border e-commerce logistics, specifically focusing on its role in achieving competitive advantage for SMEs. The research employs a comprehensive approach, considering various factors that influence the formulation and implementation of cross-border e-commerce logistics strategies. Factors such as scale, quality, potential, and infrastructure are scrutinized to provide a nuanced understanding of the dynamic interplay. Real-world data is analyzed using advanced statistical techniques to derive meaningful insights. The data of CBEC and logistics planning in Guangdong from 2013 to 2018 was used as a case example to demonstrate its and flexibility and usefulness of empirically proposed model estimation and develop a holistic indicator system for the evaluation of the synergistic development of CBEC and logistics. This research focus on the modelling of the CBEC for the synergistic effect analysis and evaluate the usefulness and flexibility of CBEC to achieve competitive advantage. The study aims to uncover insights into the most effective approaches for Chinese SMEs in navigating the landscape of cross-border e-commerce. By examining the results derived from the empirical model, the research sheds light on the impact of cross-border e-commerce logistics on competitive advantage. The findings contribute to a deeper understanding of how SMEs can strategically position themselves in the cross-border e-commerce arena. Through the analysis of real-world data and the application of advanced statistical methods, this research offers valuable insights for Chinese SMEs. The insights generated from the study not only illuminate the intricacies of cross-border e-commerce logistics but also provide practical recommendations to reduce associated costs. Ultimately, the study aims to equip SMEs with the knowledge necessary to thrive in the evolving landscape of cross-border e-commerce. [ABSTRACT FROM AUTHOR]

In recent years, the rapid development of the Internet has promoted the continuous expansion of the scale of China's tourism industry, and the amount of tourism data has surged. However, tourists need help bringing personal interest and high-value data from the plethora of tourism information. The rise of artificial intelligence has transformed traditional tourism into an intelligent, data-driven industry. This shift has generated vast tourism data, offering both opportunities and challenges. The paper discusses an AI and IoT-based Intelligent Tourism Recommendation System (ITRS) that offers travelers predefined itineraries, personalized suggestions, and tourism insights. This system simplifies attraction discovery, unveiling hidden value within tourism data at the intersection of AI and IoT. The present study thoroughly investigates AI-based recommendation algorithms before delving into the system's architecture. It categorizes user-based, project-based, and article-based collaborative filtering methodologies tailored to specific goals. First, thoroughly examine AI-based recommendation algorithms before delving into the system architecture. Second, categorize collaborative filtering methods as user-based, project-based, and article-based, each tailored to specific objectives. Third, delve into the Apriori algorithm's complexity within the context of weighted association rules and introduce an enhanced iteration for improved efficiency. The proposed scheme encompasses an elaborate ITRS plan featuring a user interest model and a client module, crucial for the computation and analysis of users' long-term and short-term interests. Rigorous performance testing confirms the ITRS's superiority across varying support levels, with experimental results demonstrating the Apriori algorithm's exceptional accuracy, achieving a 94.3% improvement over other methods. The Apriori algorithm is better than traditional recommendation algorithms such as Linear Regression, Logistic Regression, Decision Trees, Random Forest, Support Vector Machines, K-nearest neighbor, Naive Bayes, and XGBoost. [ABSTRACT FROM AUTHOR]

This paper develops a cost-effective sequential route recommender system to provide real-time routing recommendations for vacant taxis searching for the next passenger. We propose a prediction-and-optimization framework to recommend the searching route that maximizes the expected profit of the next successful passenger pickup based on the dynamic taxi demand-supply distribution. Specifically, this system features a deep learning-based predictor that dynamically predicts the passenger pickup probability on a road segment and a recursive searching algorithm that recommends the optimal searching route. The predictor integrates a graph convolution network (GCN) to capture the spatial distribution and a long short-term memory (LSTM) to capture the temporal dynamics of taxi demand and supply. The GCN-LSTM model can accurately predict the pickup probability on a road segment with the consideration of potential taxi oversupply. Then, the dynamic distribution of pickup probability is fed into the route optimization algorithm to recommend the optimal searching routes sequentially as route inquiries emerge in the system. The recursion tree-based route optimization algorithm can significantly reduce the computational time and provide the optimal routes within seconds for real-time implementation. Finally, extensive experiments using Beijing Taxi GPS data demonstrate the effectiveness and efficiency of the proposed recommender system. History: Accepted by Ram Ramesh, Area Editor for Data Science and Machine Learning. Funding: This work was partially supported by the Hong Kong Research Grants Council [Grants CityU 21500220, CityU 11504322] and the National Natural Science Foundation of China [Grant 72201222]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information (https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2021.0112) as well as from the IJOC GitHub software repository (https://github.com/INFORMSJoC/2021.0112). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/. [ABSTRACT FROM AUTHOR]

Personalized points-of-interest (POI) recommendation is very important for improving the service quality of location-based social network applications. It has become one of the most popular research directions in the industry and academia. However, the realization of high-quality personalized POI recommendation faces three major challenges: (i) the interest drift issue caused by the spatiotemporal dynamics of user check-in behavior, (ii) how to integrate as much heterogeneous information as possible to alleviate data sparseness and cold start issues, and (iii) how to use implicit feedback to model complex high-order nonlinear user-POI interactions. To jointly address all these challenges, we propose the TGVx recommendation framework and establish a dynamic personalized POI deep recommendation model, where T and G respectively represent time and geographic factors, V represents out-of-town visitors, and x represents time slot number. TGVx is composed of x parallel TGV models where the TG module mines high-order nonlinear user-POI interaction relationships and integrates multisource heterogeneous information, and the V module transfers the check-in records of out-of-town visitors in hometowns and generates pseudo check-in records in the target city. Technically, we design a new unsupervised deep learning network T-SemiDAE for the TG module. We built a POI-word heterogeneous network for the V module and used graph embedding technology to match the most similar POIs across cities and transfer check-in records. The experimental results on the actual datasets show that the TGVx model is always better than other advanced models in terms of accuracy and diversity for local and out-of-town recommendation scenarios. Compared with the best baseline model semi-deep auto-encoder with a conditional layer the average improvement rates of accuracy and diversity of TGVx are 17.1% to 58.6% and 2.25% to 28.86%, respectively. In theory, our research effectively uses data science and analysis methods to design a recommender system. In practice, our research is motivated by practical problems, and the research results have high practical promotion value. History: Accepted by Ram Ramesh, Area Editor for Data Science & Machine Learning. Funding: This work was supported by the National Natural Science Foundation of China [Grants 71831003, 72293563, and 72172025], the Dalian Science and Technology Talent Innovation Support Policy Project [The High-level Talent Innovation Team Project 2022RG17], and the Liaoning Provincial Department of Education [Basic Scientific Research Project LJKMZ20221582]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information (https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.1286) as well as from the IJOC GitHub software repository (https://github.com/INFORMSJoC/2021.0010) at (http://dx.doi.org/10.5281/zenodo.7407123). [ABSTRACT FROM AUTHOR]

Landslides with step‐like deformation features are widely distributed in the Three Gorges Reservoir area (TGR) of China, posing a severe hazard to the inhabitants of this region. This paper proposes a multi‐input and multi‐output intelligent integrated displacement prediction model for landslides with step‐like displacement patterns. In this new model, three interconnected and information‐transmitted functional sub‐models are integrated. Unsupervised learning is used to identify different landslide deformation states automatically, and the imbalance classification and explainable artificial intelligence techniques are introduced for qualitative prediction and information filtering. Probability theory and deep machine learning are adopted to provide deterministically predicted values and quantify their uncertainty. The case study of the Baijiabao landslide in the TGR region proves that the proposed model performs satisfactorily in both point and interval predictions. The intelligent integrated model can also provide the forecast of landslide deformation states, visual input information filtering and back analysis of influencing factors, which are valuable to landslide early warning and risk management. [ABSTRACT FROM AUTHOR]

The emotion analysis of hotel online reviews is discussed by using the neural network model BERT, which proves that this method can not only help hotel network platforms fully understand customer needs but also help customers find suitable hotels according to their needs and affordability and help hotel recommendations be more intelligent. Therefore, using the pretraining BERT model, a number of emotion analytical experiments were carried out through fine-tuning, and a model with high classification accuracy was obtained by frequently adjusting the parameters during the experiment. The BERT layer was taken as a word vector layer, and the input text sequence was used as the input to the BERT layer for vector transformation. The output vectors of BERT passed through the corresponding neural network and were then classified by the softmax activation function. ERNIE is an enhancement of the BERT layer. Both models can lead to good classification results, but the latter performs better. ERNIE exhibits stronger classification and stability than BERT, which provides a promising research direction for the field of tourism and hotels. [ABSTRACT FROM AUTHOR]

Foliar nutrient diagnosis is a potent tool for nutrient management in crops. Litchi (Litchi chinensis Sonn.) plantation in China dominates the worldwide litchi cultivation, and Feizixiao is the most widely cultivated species in China. Herein, 538 leaf samples were collected at fruit swelling stage and the terminal shoot maturing stage during 2016–2017, and the fruit yields of all the sampled trees were recorded. The suitability of four nutrient diagnosis methods including critical value approach (CVA), sufficiency range approach (SRA), modified diagnosis and recommendation integrated system (M-DRIS), and compositional nutrient diagnosis (CND) in Feizixiao were compared. Only a few of foliar nutrients could be diagnosed by CVA. Similar evaluations were achieved by both M-DRIS and CND, but the diagnosis accuracy highly varied with the severity of nutrient deficiency or abundance and the diagnosis year. Finally, SRA was accepted to establish the diagnosis norms due to its stable assessments for all nutrients. The optimal foliar nutrient range for Feizixiao is recommended as N 16.7–19.2, P 1.06–1.25, K 5.1–6.7, Ca 7.7–11.0, Mg 2.5–3.7 and S 1.51–1.81 g kg−1, and Zn 19.6–32.6 and B 11.5–19.2 mg kg−1 at the fruit swelling stage , and N 19.7–22.0, P 1.69–1.95, K 10.8–12.7, Ca 3.0–4.1, Mg 2.5–2.9 and S 1.38–1.57 g kg−1, and Zn 15.0–18.9 and B 10.8–16.8 mg kg−1 at the terminal shoot maturing stage. [ABSTRACT FROM AUTHOR]

In many congested areas, shared parking has gotten increasing attention because of its potential to alleviate parking resource shortages. However, managing parking resources remains a challenge when simultaneously considering multiple decision-making criteria of public travelers in allocating parking spaces and recommending optimal parking routes. To fill this gap, from four perspectives, i.e., driving, among shared parking lots, at a shared parking lot, between shared parking spaces and destinations, we proposed nine criteria for shared parking space allocations and parking route recommendations, and we also gave the quantitative models for different criteria. Furthermore, an analytic hierarchy process Entropy-TOPSIS grey relational analysis (AHP-Entropy-TOPSIS-GRA) method and an improved ant colony algorithm were proposed to solve the proposed allocation of parking spaces and recommend optimal parking routes, respectively. Finally, the validity of our proposed models and algorithms was tested by empirical parking data and road traffic data collected in Huai'an City, Jiangsu province, China. The research helps provide a theoretical foundation for implementing shared parking initiatives and improving public travelers' parking satisfaction. [ABSTRACT FROM AUTHOR]

With the rapid development of China's economy, people are eager for an effective way to relieve work pressure and strengthen their health at the same time. Outdoor sport is one of the best choices for people. However, the amount of recommended data on the network is very large. As a result, when people understand outdoor sports through the network, they cannot effectively obtain the information they want. This is the problem of "information overload," and personalized recommendation system can effectively alleviate this problem. In order to effectively recommend outdoor sports to users, a useful attempt was made in the personalized recommendation system for outdoor sports in this paper. The specific work of this paper is as follows: firstly, the current situation of outdoor sports in China was summarized, and the related technologies of the recommendation system were studied, including user modeling technology, recommendation target modeling technology, and recommendation algorithm. In order to obtain better recommendation effect, this paper proposes to mix user-based collaborative filtering recommendation algorithm, project-based collaborative filtering recommendation algorithm, and content-based recommendation algorithm. The hybrid algorithm adopts the way of feature expansion and weighted combination. Firstly, the hybrid model (model 1) of user-based collaborative filtering recommendation and content-based recommendation is obtained. Secondly, the hybrid model (model 2) based on project collaborative filtering recommendation and content-based recommendation was obtained. Finally, model 1 and model 2 were combined together to get a hybrid model with better final recommendation effect. For the common cold start problem in the recommendation system, the system adopts content-based recommendation algorithm to solve it. [ABSTRACT FROM AUTHOR]

In the network environment, online ideological and political education is facing many challenges, the mainstream values of education have been impacted, and the backward educational model, single communication method, and solidified educational content are not conducive to the smooth progress of ideological and political education. However, at the same time, its innovation also ushers in development opportunities, rich educational resources, information platforms, diversified education subjects, strongly targeted education models, etc., which can become favorable factors for educational innovation. To this end, online educators should seize the development opportunities, grasp the good network environment, actively build a network platform for ideological and political education, create a personalized ideological and political education resource recommendation system, and reform the ideological and political education classroom, effectively playing the role of the main position of educating people. Online education breaks through the limitations of time and space, allowing students to get guidance from online teachers without leaving home, and is favored by more and more primary and secondary school students because of its unique advantages. At present, the forms of tutoring for off-campus online education in China are diverse, and there are several specific teaching modes. Whether online teaching can meet the needs of users is a problem worth studying. [ABSTRACT FROM AUTHOR]

In the huge number of online university education resources, it is difficult for learners to quickly locate the resources they need, which leads to "information trek." Traditional information recommendation methods tend to ignore the characteristics of learners, who are the main subjects of education. In order to improve the recommendation accuracy, a recommendation algorithm based on improved collaborative filtering model is proposed in this paper. Firstly, according to the student behavior data, consider the behavior order to create the behavior graph and behavior route. Then, the path of text type is vectorized by the Keras Tokenizer method. Finally, the similarity between multidimensional behavior path vectors is calculated, and path collaborative filtering recommendations are performed for each dimension separately. The MOOC data of a university in China are introduced to experimentally compare the algorithm of the article as well as the control group algorithm. The results show that the proposed algorithm takes better values in evaluation indexes, thus verifying that this algorithm can improve the effectiveness of innovation and entrepreneurship education resources recommendation in universities. [ABSTRACT FROM AUTHOR]

Concentration of power in terms of user traffic and copyright content is most evident in content platforms in China. Such concentration has generated an unexpected impact on the way we understand and appreciate creativity, on copyright enforcement and determination of liability on content platforms, and on the regulation of the cultural market by the government. Specifically, the concentration of power in content platforms has not only curbed direct online piracies to a large extent but has also accelerated the fragmentation of copyright enforcement and spawned the need for algorithmic recommendation and filtering systems, which in turn has reinforced the cultural censorship system of China. This paper argues that the employment of algorithms by platforms must be treated with prudence: the algorithmic decision-making systems employed by platforms must be transparent as much as possible, and remedies must be provided for concerned users. The algorithms employed by content platforms must be adjusted to reflect not just the interests of the platforms but also the public interest in accessing and delivering information and local policy considerations. This paper suggests that our regulatory framework should reflect the algorithmic turn of content platforms in its legal and non-legal instruments and alleviate their negative impact on society. [ABSTRACT FROM AUTHOR]

Grey system theory is an effective mathematical method for studying small sample data and solving poor information problems. The grey correlation model and grey prediction model in this theory have been widely used in scientific studies in various industries. Currently, it is quite difficult for China to formulate scientific policies as livestreaming e-commerce is an emerging industry with little available annual data; therefore, the use of grey system theory is crucial to the study of the future scale of livestreaming e-commerce. In this paper, of many factors influencing the development of livestreaming e-commerce, 14 predictors of livestreaming e-commerce development scale were selected to construct a grey correlation model, by which 5 main predictors were determined. Based on the predictors identified above, 4 grey prediction models of GM (1,1), DGM (1,1), NDGM (1,1), and FDGM (1,1) were constructed, and the accuracy of these models was compared. It was concluded that the NDGM (1,1) model had the best simulation effect. The NDGM (1,1) model is then used to forecast and analyse the indicators of livestreaming e-commerce development scale from 2021 to 2023, and some relevant suggestions were made. This paper applies the new modelling approach to livestreaming e-commerce studies, thus broadening the theoretical study field of livestreaming e-commerce. Moreover, the findings can help the Chinese government make more reasonable and effective decisions as a new study on livestreaming e-commerce was conducted from a different perspective in this paper. [ABSTRACT FROM AUTHOR]

In order to understand the spatial variations of students' active school travel (AST) across different regions in mountain cities, this paper examines the spatial variations in the influence of socio-economic, topographic, and built environment characteristics on the active mode choice of students in Guiyang, a typical mountainous city in China. Using the data collected from 2021 Guiyang household travel survey, the travel characteristics of 2585 primary and secondary school students distributed in different regions within the city are analyzed. In addition, a semiparametric geographically weighted logistic regression model is employed to examine whether the determinants of students' AST vary across the studied areas. The results show that noticeable spatial variations exist in the relationship between the choice of AST and the selected explanatory variables, with different student groups exhibit sensitivity to these variables in different geographic areas. For example, among students aged 6–12 living in suburban areas, there is a higher likelihood of choosing AST as the school travel distance increases. However, students with long-distance travel in suburban areas are less inclined to choose AST as the terrain slope increases. Conversely, students aged 13–18 in urban areas are more prone to choose AST as the slope increases. To promote equitable access to AST among urban and rural students, several policy recommendations are proposed after considering their socio-economic attributes and travel environment. These recommendations primarily focus on improving school bus services, creating an active travel-friendly environment, enhancing public transit accessibility, and implementing staggered shifts for parents and students. [ABSTRACT FROM AUTHOR]

A long-term site-specific experiment of fertilizer recommendation from 2012 to 2020 was conducted to evaluate the comprehensive effects of nutrient expert decision support system on yield, benefit, fertilizer use efficiency, and soil nutrient of spring maize in northeast China. Five treatments included no-fertilizer (CK) as the control, currently traditional farmers’ practices (FP), fertilizer recommended using a nutrient expert system (NE), slow/controlled-release nitrogen fertilizer using the same rate as NE treatment (NER), and conventional fertilizer recommendation using soil testing (ST). The current study also investigated the long-term changes of maize yield, benefit, fertilizer use efficiency, the contents of inorganic N, available P, and available K in the soil, as well as the balance of nutrient input/output during the nine-year period. The results showed that the maximum, minimum, or mean values of fertilizer inputs in NE, NER, and ST treatments were all significantly lower than that of FP treatment (P<0.05), respectively. In different nutrients, NE, NER and ST treatments significantly reduced the amounts of nitrogen (N) and phosphorus (P) fertilizers application (P<0.05), but significantly increased the amount of potassium (K) fertilizer application (P<0.05), compared with that of FP treatment, respectively. Compared with FP treatment, NE, NER, and ST treatments significantly improved the stability of maize yield and net income, where increased by 5.6%-11.0% and 8.3%-14.3%, respectively. The NE treatment achieved the highest value, followed by ST and NER treatment. The recovery efficiency, agronomic efficiency, and partial factor productivity in the NE, NER, and ST treatments were significantly higher than those in FP treatment, where increased by 29.0%-40.1%, 31.3%-44.3%, and 22.0%-31.7%, respectively. Specifically, the highest value was observed in NE treatment, followed by NER and ST treatments. Compared with FP treatment, NE, NER and ST treatments significantly improved the inorganic N content in 0-30 cm soil layer (P<0.05), but significantly reduced inorganic N content in 30-90 cm soil layer (P<0.05) and available P content in 0-30 cm soil layer (P<0.05). But the soil available K content wasn’t significantly different (P>0.05) among different fertilization treatments. The contents of soil inorganic N, available P, and available K in NE treatment were close to the initial testing values before planting. The nutrient input/output balance was obtained during the nine-year period, where the N and P balances were surplus under all fertilization treatments, whereas, the K balance was deficient. But the surplus of N, P, and the deficiency of K in NE treatment attained the lowest values, followed by NER and ST treatments. In conclusion, compared with farmers’ practices and the conventional fertilizer recommendation system, the NE fertilization recommendation system fully met the nutrient requirements of maize, with optimal fertilization rate, time, and ratio. Consequently, the NE fertilization recommendation system has the potential to improve maize yield, benefits, and fertilizer use efficiency, thereby maintaining the soil nutrient stability under the reduction of fertilizer application amount, compared with farmers’ practices. However, the simplified fertilization of slow/controlled-release nitrogen fertilizer in the NE fertilization recommendation system also obtained higher benefits, but reduced labor cost. Therefore, the NE system is an appropriate fertilizer recommendation for maize in northeast China. [ABSTRACT FROM AUTHOR]

As online video service continues to grow in popularity, video content providers compete hard for more eyeball engagement. Some users visit multiple video sites to enjoy videos of their interest while some visit exclusively one site. However, due to the isolation of data, mining and exploiting user behaviors in multiple video websites remain unexplored so far. In this work, we try to model user preferences in six popular video websites with user viewing records obtained from a large ISP in China. The empirical study shows that users exhibit both consistent cross-site interests as well as site-specific interests. To represent this dichotomous pattern of user preferences, we propose a generative model of Multi-site Probabilistic Factorization (MPF) to capture both the cross-site as well as site-specific preferences. Besides, we discuss the design principle of our model by analyzing the sources of the observed site-specific user preferences, namely, site peculiarity and data sparsity. Through conducting extensive recommendation validation, we show that our MPF model achieves the best results compared to several other state-of-the-art factorization models with significant improvements of F-measure by 12.96, 8.24 and 6.88 percent, respectively. Our findings provide insights on the value of integrating user data from multiple sites, which stimulates collaboration between video service providers. [ABSTRACT FROM AUTHOR]

The recommendation system in the online medical consultation website is a system to assist patients to find appropriate doctors. Based on the analysis of the current situation of the development of an online medical community (Haodf.com) in China, this paper puts forward recommendation suggestions of finding the right hospital and doctor to promote the rapid integration of Internet technology and traditional medical services. A new recommendation model called Probabilistic Matrix Factorization integrated with Convolutional Neural Network (PMF-CNN) is proposed in the paper. Doctors' data in Haodf.com were used to evaluate the performance of our system. The model improves the performance of medical consultation recommendations by fusing review text and doctor information based on CNN (Convolutional Neural Network). Specifically, CNN is used to learn the feature representation of the review text and the doctors' information. Furthermore, the extended matrix factorization model is exploited to fuse the review information feature and the initial value of the doctors' information for recommendation. As is shown in the experimental results on Haodf.com datasets, the proposed PMF-CNN achieves better recommendation performances than the other state-of-the-art recommendation algorithms. And the recommendation system in an online medical website improves the utilization efficiency of doctors and the balance of public health resources allocation. [ABSTRACT FROM AUTHOR]

Accurate power load forecasting has a significant effect on a smart grid by ensuring effective supply and dispatching of power. However, electric load data generally possesses the characteristics of nonlinearity, periodicity, and seasonality. For complex electric load systems, the presence of redundant information potentially reduces the real pattern extraction for load forecasting. Bearing in mind these issues, we propose an effective forecasting model in which a feature extraction module is introduced that is combined with the variational mode decomposition (VMD) with the variational autoencoder (VAE). In this combination, VMD is utilized for decomposing complex load series and VAE is used to filter the redundant information (noises) from each decomposed series. With two real data sets from China, we demonstrate that the proposed model can achieve highly accurate predictions, as we find the mean absolute percentage error (MAPE) values for one-step-ahead prediction to be 1% (Nanjing) and 0.8% (Taixing), respectively. • A dimensionality reduction approach is proposed. • A novel combination method is designed for load forecasting. • The proposed approach can improve the accuracy in actual projects. [ABSTRACT FROM AUTHOR]

• This paper studies the competition effect of product recommendation networks. • Three network attributes, centrality, connection , and scale are measured. • Their impacts on market competition in a homogeneous product market are tested at the global-network level. • A panel data of 320,981 package tour products was used to estimate the model. • Empirical results confirm the effect of three attributes on market competition. Extant studies have focused on the sales benefits of product recommendation networks; however, the competition effects of such networks have been overlooked. Understanding the competition effects of product recommendation networks is important within an e-commerce platform because maintaining a proper level of competition is critical to the platform ecosystems. This research adopts a data-driven approach and social network analysis, elaborating upon three structural properties of recommendation networks – centrality, connection, and scale – and operationalizes them as e venness of distribution, clustering of structure, and network size at the global-network level. The competition effects of the three proposed network variables were tested based on a daily panel dataset of 320,981 package tour products from Tuniu.com , a top travel service platform in China featuring homogeneous products. Empirical results confirmed that the three proposed structural network properties increase market competition in recommendation networks of homogeneous products. This research contributes to our understanding of the impact of recommendation networks on market competition in the context of homogeneous product markets and offers practical implications of substition recommendation for platform governance. Platform operators can leverage product recommendation networks to alleviate market competition and encourage the entry of sellers with new, innovative products that contribute to the e-commerce platform ecosystems. [ABSTRACT FROM AUTHOR]

As an e-commerce feature, the recommender system can enhance the consumer shopping experience and create huge benefits for businesses. The e-tourism has become one of the largest service industries with the application and popularity of recommender systems. Many studies have confirmed that the travel product recommendation is widely different from traditional recommendations. Due to the financial and time costs, travel products are usually browsed and purchased relatively infrequently compared with other traditional products (e.g., books, movies and food). In addition, choosing the appropriate travel product will be influenced by many factors, such as departure, destination and price. To tackle this challenging problem, we propose a MV-GAN (short for M ulti- V iew G raph A ttention N etwork for travel recommendation) model. It enriches user and product semantics through both metapath-guided neighbors aggregation and multi-view fusion in heterogeneous travel product recommendation graph. In particular, we design node-level and path-level attention networks for learning user and product representations from every single view. To collaboratively integrate multiple types of relationships in different views, a view-level attention mechanism is proposed to aggregate the node representations and obtain global user and product representations. We evaluate the proposed method on a public dataset and a dataset constructed from a large tourism e-commerce website in China. Extensive experiments not only validate the effectiveness of MV-GAN, but also show its potentially good interpretability. • We propose a graph neural network-based model for the travel recommendation. • User and item representations are learned in multi-view graph embedding networks. • An attention network is designed to integrating the user and product information. [ABSTRACT FROM AUTHOR]

Travel recommendation is very critical to helping users quickly find products or services that they are interested in. The key to travel recommender systems is learning user shopping intentions, which are expressed through various supervision signals, such as the clicked products and their titles. Existing travel recommendation methods commonly infer user intentions from click behaviors on travel products. However, remarkable keywords in the product title, such as departure, destination, travel time, hotel, and transportation are paid less attention. To this end, we hypothesize that modeling click sequences and product keywords in title jointly would result in a more holistic representation of a product and towards more accurate recommendations. Thus, we propose a TRKG (short for T ravel R ecommendation with K eywords G eneration) model, which fulfills the travel recommendation and keywords generation tasks simultaneously. To generate explainable outputs, unlike most previous approaches that regard the product title as a hidden feature vector, TRKG regards keywords in the product title as an additional supervision signal. Meanwhile, TRKG integrates the long-term and short-term user preferences in the travel recommendation component and the keywords generation component. To evaluate the proposed model, we constructed datasets from a large tourism e-commerce website in China. Extensive experiments demonstrate that the proposed method yields significant improvements over state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In this study, a spatiotemporal estimation method based on Funk singular value decomposition (F-SVD) that considers the spatiotemporal correlation of rainfall is proposed to improve estimations from gauge observations. Hourly rainfall data of several flood events are selected to verify the proposed method by comparing with Inverse Distance Weighting (IDW) and Ordinary Kriging (OK) in Hanjiang basin, China. The results show that (1) F-SVD has the best performance in rainfall estimation, the larger the amount of rainfall event, the greater the improvement of F-SVD method as compared to OK and IDW; (2) through the combination/integration with F-SVD, the accuracy of IDW and OK can be greatly improved. Therefore, F-SVD can be employed as a practical method to estimate rainfall spatial distribution, which is essential data for regional hydrological modelling and water resource analysis. • A spatiotemporal estimation method based on F-SVD is proposed to estimate rainfall using gauge observation. • F-SVD is utilized to decompose the spatiotemporal matrix consisted of rainfall data. • F-SVD has higher accuracy and lower uncertainty compared to OK and IDW. • Through combination with F-SVD, the accuracy of IDW and OK can be greatly improved. • It is a practical method to process data for regional hydrological modelling. [ABSTRACT FROM AUTHOR]

• Variable data extraction and prediction were carried out with 1 km × 1 km grid as unit. • Using Stacking to integrate multiple algorithms to achieve outage prediction. • ArcGIS visualizes the results and displays the grid distribution conveniently. Prediction of user outage under typhoon disaster is of great significance for power grid disaster prevention and mitigation. Based on the idea of Stacking integration in machine learning, this paper constructs a forecasting model of user outage under typhoon disaster. It includes base learner layer and meta learner layer. In the base learner layer, random forest, adaptive boosting, extremely tree, gradient boosting decision tree, support vector machine and logistic regression are selected. Then the XGBoost algorithm is selected in the meta learner layer. Taking one of the most frequently struck area Xuwen County of Guangdong, China as the research object, the model is verified by typhoon "Rammasun(2014)","Kalmaegi(2014)" and "Mujigae(2015)". The results show that the accuracy and recall of the prediction model based on multi-algorithm Stacking integration can reach 0.7678 and 0.9059 respectively. It can well realize the prediction of user outage under typhoon disaster. [ABSTRACT FROM AUTHOR]

Planning travel to unfamiliar regions is a difficult task for novice travelers. The burden can be eased if the resident of the area offers to help. In this paper, we propose a social itinerary recommendation by learning from multiple user-generated digital trails, such as GPS trajectories of residents and travel experts. In order to recommend satisfying itinerary to users, we present an itinerary model in terms of attributes extracted from user-generated GPS trajectories. On top of this itinerary model, we present a social itinerary recommendation framework to find and rank itinerary candidates. We evaluated the efficiency of our recommendation method against baseline algorithms with a large set of user-generated GPS trajectories collected from Beijing, China. First, systematically generated user queries are used to compare the recommendation performance in the algorithmic level. Second, a user study involving current residents of Beijing is conducted to compare user perception and satisfaction on the recommended itinerary. Third, we compare mobile-only approach with Mobile+Cloud architecture for practical mobile recommender deployment. Lastly, we discuss personalization and adaptation factors in social itinerary recommendation throughout the paper. [ABSTRACT FROM AUTHOR]

With rapid development of wind power in China, it has become an integral part of the energy structure, so it is of practical significance to forecast wind power generation accurately. As wind power generation in China experiences both an exponential increase trend and a seasonal fluctuation pattern, it cannot be accurately forecasted by traditional models. As the grey model GM(1,1) can capture an exponential growth trend and the Hodrick-Prescott filter is known for its capability of characterizing seasonality factors, this paper proposes a novel seasonal forecasting method that integrates the HP filter into the grey model GM(1,1). The proposed model is then applied to carry out an empirical analysis based on the seasonal wind power generation data between 2013 and 2019 in China. The forecasting results from the new model are then compared with three existing approaches. The comparison results indicate that the proposed model generally outperforms existing methods as it can well capture seasonal fluctuations in the data series. A further prediction of wind power generation in China is conducted into a future horizon of 2020 and 2021 by using our model, followed by a set of policy recommendations for further development of the wind power industry in China. • An improved seasonal grey model based on Hodrick-Prescott filter is proposed. • The proposed model could predict series with dual property of increase trend and seasonal fluctuation simultaneously. • The proposed model performs better in prediction accuracy than other traditional models. • The wind power generation of China is predicted via the model. [ABSTRACT FROM AUTHOR]

Decreasing nutrient losses from excessive synthetic fertilizer inputs is the direct and valid way to address low nutrient use efficiency and the related environmental consequences. Here, we established a comprehensive database of nitrogen (N), phosphorus (P), and carbon (C) losses from rice paddy fields in China, which we used to evaluate fertilization-induced losses and the impact of environmental factors, and to mitigate losses by adopting alternative fertilization options and setting input thresholds. Our results showed that most N-loss pathways had exponential increases with additional N input. In average, 23.8% of the N applied was lost via NH 3 (16.1%), N 2 O (0.3%), leaching (4.8%), and runoff (2.6%). Total P loss was approximately 2.7% of the input, composed of leaching (1.3%) and runoff (1.4%). C lost as CH 4 accounted for 4.9% of the organic C input. A relative importance analysis indicated that climate or soil variation rather than fertilizer rate was the dominant factor driving N and P leaching, and CH 4 emissions. Based on the sensitivity of multiple N-loss pathways to N fertilization, we propose upper thresholds for N inputs of 142–191 kg N ha−1 across four rice types, which would avoid dramatic increases in N losses. Compared to conventional chemical fertilization, alternative fertilization options had diverse performances: enhanced-efficiency N fertilizer reduced N loss rate by 7.8 percent points and the global warming potential (GWP, considering N 2 O and CH 4 emissions) by 28.8%; combined manure and chemical N fertilizer reduced N loss rate by 9.0 percent points but increased the GWP by 56.9%; straw return had no effect on total N loss but almost doubled the GWP. Using nutrient sources most appropriate to site-specific conditions is demonstrated as a robust way to decrease nutrient losses. Setting nutrient input thresholds would also contribute to the mitigation of environmental pollution, especially in regions with poor fertilization recommendation systems. Image 1 • N loss rate was 23.8% via NH 3 , N 2 O, leaching, and runoff, total P loss rate was 2.7%, and C loss rate was 4.9% via CH 4. • We proposed a method to determine the threshold of N application rate that ranged 142–191 kg N ha−1 across four rice types. • EEF N reduced N loss and the GWP, M N +CF N only reduced N loss, and SR had no effect on N loss but doubled the GWP. [ABSTRACT FROM AUTHOR]

• This research has proposed a precision agriculture technology adoption prediction model. • Need-technology fit, perceived benefit, facilitating conditions and perceived risk have significant impacts on farmers' intention to adopt precision agriculture. • Facilitating conditions (e.g. knowledge, resources and access to consultant services) were found to play the biggest role in improving the willingness of arable farmers in the North China Plain to adopt precision agriculture. • Social influence, was not found to play a significant impact on farmers' willingness to adopt precision agriculture in North China Plain agriculture. Precision agriculture has the potential to deliver improved and more sustainable food production. Despite the various Chinese policy initiatives to strengthen national food security, there is evidence that the adoption of precision agriculture technologies in China has been much lower when compared to other developed agricultural economies. This study therefore aims to explore factors that determine Chinese farmers' adoption of precision agriculture technologies in cropping systems and to provide recommendations on technology promotion in the future. The current status of precision agriculture adoption by smallholder farmers within crop farming systems in the North China Plain was explored. An integrated model of "Adapted Unified Theory of Acceptance and Usage of Technology (AUT2)" was developed to explain individual farmers' intention to adopt precision agriculture. 456 surveys were conducted via face to face interviews in the North China Plain and structural equation modelling analysis was used to estimate the proposed AUT2 model. The results showed that perceived need for technology characteristics (PNTC), perceived benefits, perception of the efficacy of facilitating conditions and perceived risks of adoption have significant impacts on farmers' intention to adopt precision agriculture. The facilitating conditions (e.g. knowledge, resources and access to consultant services) were the best predictor improving Chinese farmers' willingness to adopt these technologies. Policy makers and service providers need to consider these factors in the promotion of technologies. [ABSTRACT FROM AUTHOR]

China officially implemented the renewable portfolio standard (RPS) on 1 January 2019, and it remains uncertain as to whether this can effectively solve the problem of renewable energy consumption in China and ease the pressure of government subsidies. In order to study the impact of this policy on China's renewable energy power generation and explore RPS policy that is more suitable for the characteristics of China's renewable energy, we first develop a revenue function model based on the just released RPS policy to explore the effectiveness of the policy, the feasibility conditions for successful implementation, and the problems that may be encountered during the implementation process. Then, we propose policy recommendations based on the possible problems of the current policy and design an "incremental electricity price" supplementary policy to improve the possibility of successful implementation of the RPS policy. Finally, an evolutionary game model is established to simulate and verify the possibility of successful implementation of the supplementary policy. The main research results are: (1) the essence of the current RPS policy is the comprehensive implementation policy of the RPS and feed-in-tariff (FiT); (2) because of the characteristics of China's energy structure, the implementation of this policy reform is more resistant; (3) the quantitative research on the revenue function model shows that the current transaction price of the green certificate market is very low, which is not conducive to alleviating the state's subsidy pressure on renewable energy power generation; and (4) analysis of empirical data shows that the successful implementation of the "incremental electricity price" policy relies on the initial strategies of grid companies and users. [ABSTRACT FROM AUTHOR]