Showing papers by "Zhejiang Gongshang University published in 2021"

Ecological effects of new-type urbanization in China

Abstract: The development of urbanization in China has changed from a traditional form of urbanization that focuses on the rate of growth to a new type of urbanization that stresses improvements in quality. To evaluate this new type of urbanization, this paper constructs a system that looks at the four dimensions of economy, population, society, and environment, and then, using provincial-level panel data, employs a dynamic spatial panel model to empirically test the ecological effects of the new type urbanization. The study finds that the new-type urbanization in China increased gradually from 2003 to 2017, focused on improvements to the ecological environment, and displayed obvious inter-provincial differences. Moreover, China's new-type urbanization has not only effectively reduced pollution emissions and improved energy efficiency but has also been significant in terms of its ecological effect. Moreover, economic urbanization, population urbanization, social urbanization and environmental urbanization exhibit the obvious ecological effects of “pollution reduction and efficiency improvement.” In the process of this new type of urbanization, both the government's “severe constraints” on pollution emissions and the active introduction of foreign capital are further important avenues that lead to achieving “pollution reduction and efficiency improvements.”

Learning Intents behind Interactions with Knowledge Graph for Recommendation

Abstract: Knowledge graph (KG) plays an increasingly important role in recommender systems. A recent technical trend is to develop end-to-end models founded on graph neural networks (GNNs). However, existing GNN-based models are coarse-grained in relational modeling, failing to (1) identify user-item relation at a fine-grained level of intents, and (2) exploit relation dependencies to preserve the semantics of long-range connectivity. In this study, we explore intents behind a user-item interaction by using auxiliary item knowledge, and propose a new model, Knowledge Graph-based Intent Network (KGIN). Technically, we model each intent as an attentive combination of KG relations, encouraging the independence of different intents for better model capability and interpretability. Furthermore, we devise a new information aggregation scheme for GNN, which recursively integrates the relation sequences of long-range connectivity (i.e., relational paths). This scheme allows us to distill useful information about user intents and encode them into the representations of users and items. Experimental results on three benchmark datasets show that, KGIN achieves significant improvements over the state-of-the-art methods like KGAT, KGNN-LS, and CKAN. Further analyses show that KGIN offers interpretable explanations for predictions by identifying influential intents and relational paths. The implementations are available at this https URL.

Learning Intents behind Interactions with Knowledge Graph for Recommendation

Abstract: Knowledge graph (KG) plays an increasingly important role in recommender systems. A recent technical trend is to develop end-to-end models founded on graph neural networks (GNNs). However, existing GNN-based models are coarse-grained in relational modeling, failing to (1) identify user-item relation at a fine-grained level of intents, and (2) exploit relation dependencies to preserve the semantics of long-range connectivity. In this study, we explore intents behind a user-item interaction by using auxiliary item knowledge, and propose a new model, Knowledge Graph-based Intent Network (KGIN). Technically, we model each intent as an attentive combination of KG relations, encouraging the independence of different intents for better model capability and interpretability. Furthermore, we devise a new information aggregation scheme for GNN, which recursively integrates the relation sequences of long-range connectivity (i.e., relational paths). This scheme allows us to distill useful information about user intents and encode them into the representations of users and items. Experimental results on three benchmark datasets show that, KGIN achieves significant improvements over the state-of-the-art methods like KGAT [41], KGNN-LS [38], and CKAN [47]. Further analyses show that KGIN offers interpretable explanations for predictions by identifying influential intents and relational paths. The implementations are available at https://github.com/huangtinglin/Knowledge_Graph_based_Intent_Network.

Dual Encoding for Video Retrieval by Text.

Abstract: This paper attacks the challenging problem of video retrieval by text. In such a retrieval paradigm, an end user searches for unlabeled videos by ad-hoc queries described exclusively in the form of a natural-language sentence, with no visual example provided. Given videos as sequences of frames and queries as sequences of words, an effective sequence-to-sequence cross-modal matching is crucial. To that end, the two modalities need to be first encoded into real-valued vectors and then projected into a common space. In this paper we achieve this by proposing a dual deep encoding network that encodes videos and queries into powerful dense representations of their own. Our novelty is two-fold. First, different from prior art that resorts to a specific single-level encoder, the proposed network performs multi-level encoding that represents the rich content of both modalities in a coarse-to-fine fashion. Second, different from a conventional common space learning algorithm which is either concept based or latent space based, we introduce hybrid space learning which combines the high performance of the latent space and the good interpretability of the concept space. Dual encoding is conceptually simple, practically effective and end-to-end trained with hybrid space learning. Extensive experiments on four challenging video datasets show the viability of the new method.

Smart and Practical Privacy-Preserving Data Aggregation for Fog-Based Smart Grids

Abstract: With the increasingly powerful and extensive deployment of edge devices, edge/fog computing enables customers to manage and analyze data locally, and extends computing power and data analysis applications to network edges. Meanwhile, as the next generation of the power grid, the smart grid can achieve the goal of efficiency, economy, security, reliability, use safety and environmental friendliness for the power grid. However, privacy and secure issues in fog-based smart grid communications are challenging. Without proper protection, customers’ privacy will be readily violated. This article presents a smart and practical Privacy-preserving Data Aggregation (PDA) scheme with smart pricing and packing method for fog-based smart grids, which achieves diversified tariffs, multifunctional statistics and efficiency. Especially, we first propose a smart PDA scheme with Smart Pricing (PDA-SP). With PDA-SP, the Control Center (CC) can compute more complex and higher-order aggregation statistics to provide various services, provide diversiform pricing strategies and choose a double-winning strategy. Subsequently, we put forward a practical PDA scheme with Packing Method (PDA-PM), which is able to reduce the size of encrypted data and improve performance in performing various secure computations. Moreover, we extend our original packing method and present a more useful packing method, which can handle general vectors with large entries. The security analysis shows that our proposed scheme is secure against many threats. The performance evaluation reveals that the computation and communication overheads of our proposed scheme are effectively reduced by employing the Somewhat Homomorphic Encryption (SHE), and our packing method can further significantly reduce these overheads.

VAE-Stega: Linguistic Steganography Based on Variational Auto-Encoder

Abstract: In recent years, linguistic steganography based on text auto-generation technology has been greatly developed, which is considered to be a very promising but also a very challenging research topic. Previous works mainly focus on optimizing the language model and conditional probability coding methods, aiming at generating steganographic sentences with better quality. In this paper, we first report some of our latest experimental findings, which seem to indicate that the quality of the generated steganographic text cannot fully guarantee its steganographic security, and even has a prominent perceptual-imperceptibility and statistical-imperceptibility conflict effect (Psic Effect). To further improve the imperceptibility and security of generated steganographic texts, in this paper, we propose a new linguistic steganography based on Variational Auto-Encoder (VAE), which can be called VAE-Stega. We use the encoder in VAE-Stega to learn the overall statistical distribution characteristics of a large number of normal texts, and then use the decoder in VAE-Stega to generate steganographic sentences which conform to both of the statistical language model as well as the overall statistical distribution of normal sentences, so as to guarantee both the perceptual-imperceptibility and statistical-imperceptibility of the generated steganographic texts at the same time. We design several experiments to test the proposed method. Experimental results show that the proposed model can greatly improve the imperceptibility of the generated steganographic sentences and thus achieves the state of the art performance.

Context-aware Biaffine Localizing Network for Temporal Sentence Grounding

Abstract: This paper addresses the problem of temporal sentence grounding (TSG), which aims to identify the temporal boundary of a specific segment from an untrimmed video by a sentence query. Previous works either compare pre-defined candidate segments with the query and select the best one by ranking, or directly regress the boundary timestamps of the target segment. In this paper, we propose a novel localization framework that scores all pairs of start and end indices within the video simultaneously with a biaffine mechanism. In particular, we present a Context-aware Biaffine Localizing Network (CBLN) which incorporates both local and global contexts into features of each start/end position for biaffine-based localization. The local contexts from the adjacent frames help distinguish the visually similar appearance, and the global contexts from the entire video contribute to reasoning the temporal relation. Besides, we also develop a multi-modal self-attention module to provide fine-grained query-guided video representation for this biaffine strategy. Extensive experiments show that our CBLN significantly outperforms state-of-thearts on three public datasets (ActivityNet Captions, TACoS, and Charades-STA), demonstrating the effectiveness of the proposed localization framework. The code is available at https://github.com/liudaizong/CBLN.

Combining Graph Neural Networks with Expert Knowledge for Smart Contract Vulnerability Detection

Abstract: Smart contract vulnerability detection draws extensive attention in recent years due to the substantial losses caused by hacker attacks. Existing efforts for contract security analysis heavily rely on rigid rules defined by experts, which are labor-intensive and non-scalable. More importantly, expert-defined rules tend to be error-prone and suffer the inherent risk of being cheated by crafty attackers. Recent researches focus on the symbolic execution and formal analysis of smart contracts for vulnerability detection, yet to achieve a precise and scalable solution. Although several methods have been proposed to detect vulnerabilities in smart contracts, there is still a lack of effort that considers combining expert-defined security patterns with deep neural networks. In this paper, we explore using graph neural networks and expert knowledge for smart contract vulnerability detection. Specifically, we cast the rich control- and data- flow semantics of the source code into a contract graph. To highlight the critical nodes in the graph, we further design a node elimination phase to normalize the graph. Then, we propose a novel temporal message propagation network to extract the graph feature from the normalized graph, and combine the graph feature with designed expert patterns to yield a final detection system. Extensive experiments are conducted on all the smart contracts that have source code in Ethereum and VNT Chain platforms. Empirical results show significant accuracy improvements over the state-of-the-art methods on three types of vulnerabilities, where the detection accuracy of our method reaches 89.15%, 89.02%, and 83.21% for reentrancy, timestamp dependence, and infinite loop vulnerabilities, respectively.

Production, bioactive properties, and potential applications of fish protein hydrolysates: Developments and challenges

Abstract: Background The health benefits associated with consuming fish products are mainly attributed to their desirable nutrition profiles, including vitamins, minerals, essential amino acids, and polyunsaturated fatty acids. However, large quantities of fish proteins are presently underutilized or discarded as waste. Effective strategies to utilize fish proteins are therefore needed. Recently, researchers have focused on generating and characterizing bioactive fish protein hydrolysates and peptides and then studying their potential health benefits. Scope and Approach The major methods of producing, separating, and purifying protein hydrolysates are initially given. Then, the biological activities and potential mechanisms of action of protein hydrolysates and peptides are discussed. Finally, current limitations and future possibilities of fish peptide identification, production, and bioactivity are identified and discussed. Key findings and conclusions Fermentation, chemical synthesis, and enzymatic hydrolysis are effective methods of obtaining hydrolysates from underutilized fish protein by-products. These hydrolysates can then be purified by membrane separation and chromatographic methods to obtain bioactive peptides. The molecular characteristics of the peptides can then be identified using mass spectrometry. Fish hydrolysates/peptides have multiple biological activities, including antioxidative, lipid homeostasis modulation, anti-inflammatory, anticancer, neuroprotective, and antihypertensive activities, which make them promising nutraceutical ingredients for application in foods. Moreover, they often have emulsifying, foaming, and gelling properties, which means they may be suitable as multipurpose functional ingredients. Thus, waste-derived fish by-products may be turned into value-added functional ingredients designed to address chronic diseases. However, further research is required to develop large-scale commercially viable extraction and purification methods, develop robust structure-function relationships for peptides, and perform in vivo human studies of peptide bioactivity.

Unfolding the Determinants of COVID-19 Vaccine Acceptance in China.

Abstract: Background: China is at the forefront of global efforts to develop COVID-19 vaccines and has five fast-tracked candidates at the final-stage, large-scale human clinical trials testing phase. Vaccine-promoting policymaking for public engagement is a prerequisite for social mobilization. However, making an informed and judicious choice is a dilemma for the Chinese government in the vaccine promotion context. Objective: In this study, public opinions in China were analyzed via dialogues on Chinese social media, based on which Chinese netizens’ views on COVID-19 vaccines and vaccination were investigated. We also aimed to develop strategies for promoting vaccination programs in China based on an in-depth understanding of the challenges in risk communication and social mobilization. Methods: We proposed a novel behavioral dynamics model, SRS/I (susceptible-reading-susceptible/immune), to analyze opinion transmission paradigms on Chinese social media. Coupled with a meta-analysis and natural language processing techniques, the emotion polarity of individual opinions was examined in their given context. Results: We collected more than 1.75 million Weibo messages about COVID-19 vaccines from January to October 2020. According to the public opinion reproduction ratio (R0), the dynamic propagation of those messages can be classified into three periods: the ferment period (R01=1.1360), the revolution period (R02=2.8278), and the transmission period (R03=3.0729). Topics on COVID-19 vaccine acceptance in China include price and side effects. From September to October, Weibo users claimed that the vaccine was overpriced, making up 18.3% (n=899) of messages; 38.1% (n=81,909) of relevant topics on Weibo received likes. On the contrary, the number of messages that considered the vaccine to be reasonably priced was twice as high but received fewer likes, accounting for 25.0% (n=53,693). In addition, we obtained 441 (47.7%) positive and 295 (31.9%) negative Weibo messages about side effects. Interestingly, inactivated vaccines instigated more heated discussions than any other vaccine type. The discussions, forwards, comments, and likes associated with topics related to inactivated vaccines accounted for 53% (n=588), 42% (n=3072), 56% (n=3671), and 49% (n=17,940), respectively, of the total activity associated with the five types of vaccines in China. Conclusions: Most Chinese netizens believe that the vaccine is less expensive than previously thought, while some claim they cannot afford it for their entire family. The findings demonstrate that Chinese individuals are inclined to be positive about side effects over time and are proud of China’s involvement with vaccine development. Nevertheless, they have a collective misunderstanding about inactivated vaccines, insisting that inactivated vaccines are safer than other vaccines. Reflecting on netizens’ collective responses, the unfolding determinants of COVID-19 vaccine acceptance provide illuminating benchmarks for vaccine-promoting policies.

The gut microbiota as a target to control hyperuricemia pathogenesis: Potential mechanisms and therapeutic strategies.

Abstract: Hyperuricemia (HUA) is a metabolic disorder caused by abnormal uric acid (UA) metabolism, which is a complex physiological process involving multiple organs (liver, kidney, and intestine) Although UA metabolism in the liver and kidneys has been elucidated, only a few studies have focused on the process in the intestine With our growing knowledge of the effects of gut microorganisms on health, the gut microbiota has been identified as a new target for HUA treatment In this review, the relationship between HUA and the gut microbiota is elucidated, and anti-hyperuricemia mechanisms targeting the intestine are discussed, such as the promotion of purine and UA catabolism by the gut microbiota, increases in UA excretion by the gut microbiota and its metabolites, regulation of UA absorption or secretion in the intestinal tract by certain transporters, and the intestinal inflammatory response to the gut microbiota Additionally, probiotics (Bifidobacteria and Lactobacilli) and prebiotics (polyphenols, peptides, and phytochemicals) with UA-lowering effects targeting the intestinal tract are summarized, providing reference and guidance for further research

Combining Graph Neural Networks with Expert Knowledge for Smart Contract Vulnerability Detection

Abstract: Smart contract vulnerability detection draws extensive attention in recent years due to the substantial losses caused by hacker-attacks. Existing efforts for contract security analysis heavily rely on rigid rules defined by experts, which is labor-intensive and non-scalable. More importantly, expert-defined rules tend to be error-prone and suffer the inherent risk of being cheated by crafty attackers. Recent researches focus on the symbolic execution and formal analysis of smart contract for vulnerability detection, yet to achieve a precise and scalable solution. Although several methods have been proposed to detect vulnerabilities in smart contracts, there is still a lack of effort that considers combining expert-defined security patterns with deep neural networks. In this paper, we explore using graph neural networks and expert knowledge for smart contract vulnerability detection. Specifically, we cast the rich control- and data- flow semantics of the source code into a contract graph. Then, we propose a novel temporal message propagation network to extract graph feature from the normalized graph, and combine the graph feature with expert patterns to yield a final detection system. Extensive experiments are conducted on all the smart contracts that have source code in two platforms. Empirical results show significant accuracy improvements over state-of-the-art methods.

Deep Dual Consecutive Network for Human Pose Estimation

Abstract: Multi-frame human pose estimation in complicated situations is challenging. Although state-of-the-art human joints detectors have demonstrated remarkable results for static images, their performances come short when we apply these models to video sequences. Prevalent shortcomings include the failure to handle motion blur, video defocus, or pose occlusions, arising from the inability in capturing the temporal dependency among video frames. On the other hand, directly employing conventional recurrent neural networks incurs empirical difficulties in modeling spatial contexts, especially for dealing with pose occlusions. In this paper, we propose a novel multi-frame human pose estimation framework, leveraging abundant temporal cues between video frames to facilitate keypoint detection. Three modular components are designed in our framework. A Pose Temporal Merger encodes keypoint spatiotemporal context to generate effective searching scopes while a Pose Residual Fusion module computes weighted pose residuals in dual directions. These are then processed via our Pose Correction Network for efficient refining of pose estimations. Our method ranks No.1 in the Multi-frame Person Pose Estimation Challenge on the large-scale benchmark datasets PoseTrack2017 and PoseTrack2018. We have released our code, hoping to inspire future research.

Entrepreneurship and high-quality economic development: based on the triple bottom line of sustainable development

Abstract: High-quality development has recently become an inevitable requirement for sustainable and healthy economic development in China, which pursues a sustainable ecology and a happy society for people while developing economy Entrepreneurship is believed to be the essence of enterprise and one of the important factors of social economic development This paper uses triple bottom line of sustainable development to measure high-quality economic development and studies the mutual influences between entrepreneurship and triple bottom line of sustainable development Furthermore, this paper uses three moderating variables to study the moderation effect, including business environmental index, environmental regulation and foreign direct investment The empirical results show that there exists a close relationship between entrepreneurship and triple bottom line of sustainable development There are different influences of entrepreneurship and triple bottom line of sustainable development in different areas in China The results also show that the business environment and foreign direct investment not only have direct impacts but also moderation effects on entrepreneurship towards triple bottom line of sustainable development Environmental regulation affects environmental pollution only by the moderation effect towards entrepreneurship, and the effects are different in various areas

Consensus on the terminologies and methodologies for masticatory assessment.

Abstract: A large number of methodological procedures and experimental conditions are reported to describe the masticatory process. However, similar terms are sometimes employed to describe different methodologies. Standardisation of terms is essential to allow comparisons among different studies. This article was aimed to provide a consensus concerning the terms, definitions and technical methods generally reported when evaluating masticatory function objectively and subjectively. The consensus is based on the results from discussions and consultations among world-leading researchers in the related research areas. Advantages, limitations and relevance of each method are also discussed. The present consensus provides a revised framework of standardised terms to improve the consistent use of masticatory terminology and facilitate further investigations on masticatory function analysis. In addition, this article also outlines various methods used to evaluate the masticatory process and their advantages and disadvantages in order to help researchers to design their experiments.