Effect Of Embedding Research Articles

Efficient approach for ground dynamic impedance function of the large-scale nuclear power plant footing under a layered site considering the embedded effect

Simplified lumped-parameter model in nuclear power standards have been commonly applied for ground-nuclear island structure interaction investigation. However, when considering the fine far-field radiation damping effect, the previous simplified models cannot reflect the correlation of the infinite-domain dynamic impedance function with respect to the excitation frequency as well as take into consideration the non-homogeneity of layered ground. Furthermore, the global coupling characteristics of the high-precision artificial boundaries will lead to a sudden increase in the computational cost. In this study, a large-scale calculation method for nuclear island structures through the scaled boundary finite element method (SBFEM) is developed, which can consider the embedded effect and is applicable to the coupling of the dynamic impedance function under layered ground conditions. A method of decoupling the near and far field interface domain is introduced, which solves the full matrix storage limitation for dynamic stiffness (DS) matrix according to original SBFEM, and adopts the partitioned parallel technique to improve the inverse efficiency of large matrices in the process of dynamic condensation. Numerical examples show that the proposed algorithm has good accuracy and applicability for different foundation shapes, foundation embedment effects, and layered characteristics of ground, and provides engineering references for the calculation of dynamic impedance function under large-scale nuclear power plants.

Boosting bio-lipids hydrodeoxygenation via highly dispersed and coking-resistance bimetallic Ni-La/SiO2 catalyst

A highly dispersed and coking-resistance bimetallic Ni-La/SiO2 catalyst was prepared for the selective hydrodeoxygenation (HDO) of fatty acid methyl esters (FAMEs) and non-edible bio-lipids to hydrocarbon compounds. Under optimum conditions, full conversion of methyl palmitate (MP) and 97.6% selectivity for pentadecane (PD) were achieved. The catalyst also exhibited high conversion and excellent desired product selectivity for the HDO of various model FAMEs, fatty acids, trilaurin, jatropha oil (JO) and waste cooking oil (WCO). Multiple characterizations show that the embedding effect (metal−support or metal−metal interaction) associated with the introduction of metallic La not only contributes to the formation of well-dispersed Ni particles and abundant active Niδ+−OV−LaOx species, but also prevents the migration and sintering of metallic Ni particles at high temperatures. Furthermore, the mutual transformation of the La2O3 and La2O2CO3 crystal phases effectively inhibits coke deposition on the catalyst surface during the reaction process, resulting in remarkable stability and reusability of the catalyst. A plausible reaction network and mechanism for the catalytic HDO of MP over Ni-La/SiO2 catalyst were proposed. This method of constructing a catalytic system can provide a basis for the development of efficient catalysts for the HDO of non-edible bio-lipids.

Ensemble graph auto-encoders for clustering and link prediction.

Graph auto-encoders are a crucial research area within graph neural networks, commonly employed for generating graph embeddings while minimizing errors in unsupervised learning. Traditional graph auto-encoders focus on reconstructing minimal graph data loss to encode neighborhood information for each node, yielding node embedding representations. However, existing graph auto-encoder models often overlook node representations and fail to capture contextual node information within the graph data, resulting in poor embedding effects. Accordingly, this study proposes the ensemble graph auto-encoders (E-GAE) model. It utilizes the ensemble random walk graph auto-encoder, the random walk graph auto-encoder of the ensemble network, and the graph attention auto-encoder to generate three node embedding matrices Z. Then, these techniques are combined using adaptive weights to reconstruct a new node embedding matrix. This method addresses the problem of low-quality embeddings. The model's performance is evaluated using three publicly available datasets (Cora, Citeseer, and PubMed), indicating its effectiveness through multiple experiments. It achieves up to a 2.0% improvement in the link prediction task and a 9.4% enhancement in the clustering task. Our code for this work can be found at https://github.com/xcgydfjjjderg/graphautoencoder.

Deposition behavior of the gas-atomized 7075Al and TiB2/7075Al composite powders during cold spraying

To fabricate high-quality coatings or components of high-strength 7075Al alloy by cold spraying, it is essential to understand the impact behavior and bonding mechanisms of the feedstock powder particles. For providing a general comparison, 7075Al powder and TiB2/7075Al composite powder (7075Al alloy matrix reinforced with in-situ formed and uniformly dispersed nano-TiB2 particles) produced by gas-atomization were used as feedstocks in the present study. Single particle impact tests combined with finite element analysis (FEA) were performed on the pure Al and 7075Al-T6 substrates under different process parameter sets. To provide a more realistic description, the real powder strength and plastic parameters obtained by single-particle compression tests were used as input data in the FEA model. The results show that the presence of nano-TiB2 particles has significant strengthening effects on the composite powder, thus resulting in different plastic deformation behavior upon impact compared to 7075Al powder. When the composite particles impacted the soft material (pure Al), the substrate experienced a high degree of deformation, which led to high deposition efficiency due to the embedding effect. Comparatively, the hard substrate (7075Al-T6) was less deformed, whereas the composite particle was highly deformed. Compared to the composite particle, the 7075Al particle presented a slightly higher plastic deformation and pronounced metal jets, which led to a lower critical impact velocity for successful bonding. Interestingly, a fracture was observed at the grain boundaries of both 7075Al and TiB2/7075Al composite particles in the case of a high gas temperature regime, which probably revealed that the high shock energy associated with high-velocity impact led to such intergranular fracture.

Analytical investigation on the rotational behavior of dovetail mortise–tenon joints between beams and columns in traditional Chinese timber frames

We theoretically analyzed the rotational behavior of beam–column dovetail joints in traditional Chinese timber frames in this study. An analytical model of dovetail joints at both the column head and body was designed by clarifying the moment generation mechanism and effect of rotational embedment yielding in timber perpendicular to the grain on the rotational behavior of joints. An asynchronous manifestation of rotational embedment deformation across the column surface, tenon cheeks, and upper and lower surfaces of the tenon head was analyzed, and the corresponding characteristic yield points and consequent reduction in rotational stiffness were derived in the model. The Inayama embedment theory was used to clarify the effect of rotational embedment with varying end lengths on the movement of the joint rotation center and asymmetric moment generated in different rotation directions of the column head joint. The precision of the analytical model was validated through a comparative analysis by involving nine sets of experimental data, for estimating the initial stiffness, post-yield stiffness, and identified yield points. The implications of the parameters, including the initial gap between the tenon and mortise, geometric dimensions of the dovetail tenon, and friction coefficient, were also discussed. Controlling the ratio of the initial gap and tenon height within 0.01 to ensure a certain rotational resistance of dovetail beam–column joints within the collapse limits of traditional timber frames is recommended considering the significant effect of the initial gap on the initial sliding angle and moment reduction.

Study on the Effect of Acupuncture Point Embedding on the Behavior of Rats with Allergic Rhinitis and its Mechanism of Action

To observe the interventional effect of Yingxiang acupoint embedding on the behavior of rats with allergic rhinitis and the effect on the levels of serum inflammatory factors IgE, sIgE, IL-4, and IL-5.

Preparation and characterization of Pummelo essential oil microcapsules and their application to preservation of <i>Agricus bisporus</i> (white mushrooms)

Pummelo essential oil is a natural plant oil with antioxidant and antibacterial properties; however, it is poorly stabilized and volatile, which limits its application in food preservation research. In the present work, pummelo essential oil microcapsules were prepared using the complex coalescence method, with pummelo essential oil as the core material and glutenin as the wall material. Pummelo essential oil was successfully encapsulated by glutenin, forming a stable microcapsule structure with enhanced thermal stability. Microcapsules (GH-4) with the addition of 4 mL of pummelo essential oil exhibited the best embedding effect, achieving an encapsulation efficiency of 91%. Compared with GH-0, white mushrooms preserved with GH-4 effectively maintained the hardness of the fruiting body (about 7.5 N), delayed browning (the L* value was about 80), inhibited respiration (about 18×103 mL kg-1 h-1), and slowed the quality deterioration of the white mushrooms. In conclusion, GH-4 microcapsules can be used not only for the preservation of fruits and vegetables but also provide a theoretical basis for sustainable food preservation.

The Quality of Embedding Charges Is Critical for Convergence of Many-Body Expansions When BSSE Is Absent.

There is conflicting evidence in the literature concerning the benefits of charge embedding on the convergence of many-body expansions (MBEs). Using a systematic series of water and ion-water clusters of varying size, this study indicates that the effects of charge embedding can be masked by basis set superposition error (BSSE). When BSSE is removed, this study demonstrates that charge embedding can significantly accelerate MBE convergence, where the electrostatically embedded two-body method, EE-MBE(2), can often yield accuracy close to the four-body method, MBE(4). Contrary to previous studies on smaller systems, this work shows that the performance of EE-MBE is highly sensitive to the charge model, with the best performance obtained when the natural population analysis (NPA) charge model is used and generated at the same level of theory used in the subsystem and supersystem calculations. It was demonstrated that the "3c" composite method, PBEh-3c, yields NPA atomic charges that are in excellent agreement with those obtained from supersystem density functional theory calculations. The linear-scaling X-Polarization method provides a more general approach to estimating these supersystem QM atomic charges, but its performance depends on how the fragments are defined.

Global value chain embedding and enterprise energy efficiency: A worldwide firm-level analysis

Global value chain embedding brings growth opportunities for small and medium-sized enterprises, despite their continuously controversial environmental effects. This study examines the impact of global value chain embedding on the energy efficiency of small and medium-sized enterprises using a worldwide firm-level micro data. Results support that global value chain embedding significantly improves enterprise energy efficiency, and this conclusion holds even after self-selection and endogeneity are considered. Global value chain embedding promotes enterprise innovation and labor productivity through market scale and induced learning effects and improves the energy efficiency. Moreover, the evidence only supports the lock-in effect of low value-added, while the lock-in hypothesis of pollution link and foreign technology is invalid. The positive impact of global value chain embedding on energy efficiency is greater for non-high-tech industries, mainly coming from export learning effect and innovation effect. Furthermore, there is no industry heterogeneity in the lock-in effect. Country heterogeneity in the effect of global value chain embedding on energy efficiency characteristics of regional embedding and industrialization, but not to economic development and human capital.

The Effect of Cathodic Chrome Embedding in a Zinc Coating on Protection Properties

The Effect of Cathodic Chrome Embedding in a Zinc Coating on Protection Properties

From Many-Body Ab Initio to Effective Excitonic Models: A Versatile Mapping Approach Including Environmental Embedding Effects.

We present an original multistate projective diabatization scheme based on Green's function formalisms that allows the systematic mapping of many-body ab initio calculations onto effective excitonic models. This method inherits the ability of the Bethe-Salpeter equation to describe Frenkel molecular excitons and intermolecular charge-transfer states equally well, as well as the possibility for an effective description of environmental effects in a QM/MM framework. The latter is found to be a crucial element in order to obtain accurate model parameters for condensed phases and to ensure their transferability to excitonic models for extended systems. The method is presented through a series of examples illustrating its quality, robustness, and internal consistency.

GraphNABP: Identifying nucleic acid-binding proteins with protein graphs and protein language models

The computational identification of nucleic acid-binding proteins (NABP) is of great significance for understanding the mechanisms of these biological activities and drug discovery. Although a bunch of sequence-based methods have been proposed to predict NABP and achieved promising performance, the structure information is often overlooked. On the other hand, the power of popular protein language models (pLM) has seldom been harnessed for predicting NABPs. In this study, we propose a novel framework called GraphNABP, to predict NABP by integrating sequence and predicted 3D structure information. Specifically, sequence embeddings and protein molecular graphs were first obtained from ProtT5 protein language model and predicted 3D structures, respectively. Then, graph attention (GAT) and bidirectional long short-term memory (BiLSTM) neural networks were used to enhance feature representations. Finally, a fully connected layer is used to predict NABPs. To the best of our knowledge, this is the first time to integrate AlphaFold and protein language models for the prediction of NABPs. The performances on multiple independent test sets indicate that GraphNABP outperforms other state-of-the-art methods. Our results demonstrate the effectiveness of pLM embeddings and structural information for NABP prediction. The codes and data used in this study are available at https://github.com/lixiangli01/GraphNABP.

Party organization embedding and enterprise labor income share

This study investigates how party organization embedding affects the share of labor income in enterprises. It looks at party organization embedding from two perspectives: structural embedding and relational embedding. The study finds that party organization embedding boosts the share of labor income, with enterprise greening transformation and financing constraints playing a mediating role. The study's conclusions remain valid even after robustness and endogeneity tests. Additionally, there are differences in the impact of party organization embedding on labor income share across enterprises in different regions. The effect of party organization embedding on labor income share is more significant in the western region, and party organization embedding has a more pronounced impact on labor income share in less market-oriented environments.

Cross-lingual dependency parsing for a language with a unique script

Abstract Syntactic parsing is one of the areas in Natural Language Processing. The development of large-scale multilingual language models has enabled cross-lingual parsing approaches, which allows us to develop parsers for languages that do not have treebanks available. However, these approaches rely on the assumption that languages share orthographic representations and lexical entries. In this article, we investigate methods for developing a dependency parser for Xibe, a low-resource language that is written in a unique script. We first investigate lexicalized monolingual dependency parsing experiments to examine the effectiveness of word, part-of-speech, and character embeddings as well as pre-trained language models. Results show that character embeddings can significantly improve performance, while pre-trained language models decrease performance since they do not recognize the Xibe script. We also train delexicalized monolingual models, which yield competitive results to the best lexicalized model. Since the monolingual models are trained on a very small training set, we also investigate lexicalized and delexicalized cross-lingual models. We use six closely related languages as source language, which cover a wide range of scripts. In this setting, the delexicalized models achieve higher performance than lexicalized models. A final experiment shows that we can increase performance of the cross-lingual model by combining source languages and selecting the most similar sentences to Xibe as training set. However, all cross-lingual parsing results are still considerably lower than the monolingual model. We attribute the low performance of cross-lingual methods to syntactic and annotation differences as well as to the impoverished input of Universal Dependency Part-of-Speech tags that the delexicalized model has access to.

Weighing Milky Way and Andromeda in an expanding ΛCDM Universe

The dynamics of the Local Group (LG), especially the contribution of the Milky Way (MW) and Andromeda (M 31) galaxies, is sensitive to the presence of dark energy. This work analyzes the evolution of the LG by considering it as a two-body problem in a homogeneous and isotropic expanding spacetime in a full Λcold dark matter (ΛCDM) background. Using the timing argument (TA), which links LG dynamics to LG mass, we find that the full ΛCDM background predicts a ∼10% lower mass for the LG; whereas Λ alone predicts a ∼10% higher mass. The TA mass is modified by (i) simulations and (ii) the effect of the Large Magellanic Cloud (LMC) to alleviate the poorly constrained internal mass distributions of M 31 and the MW, their time evolution, and the unknown distribution of dark matter between them. First, using IllustrisTNG simulations, we accounted for the effects of two extended halos and their environment (rather than point particles) and predicted their mass (3.89 ± 0.62)×1012 M⊙. Second, the LMC effectively changes the separation and velocities of M 31 towards the MW and reduces the predicted mass to (2.33 ± 0.72)×1012 M⊙. Despite the uncertainties around dark matter between these galaxies, the overall estimated mass is compatible with the mere sum of the MW and M 31 masses. The total mass of the TA is compatible with other estimates, such as the Hubble flow and the Virial Theorem with other dwarf galaxies. The combined result shows, for the first time, that a lower mass estimate can be obtained from the TA, with a consistent embedding and other systematic effects, and without an additional dark matter halo around the galaxies.

Contrastive prototype network with prototype augmentation for few-shot classification

In recent years, metric-based meta-learning methods have received widespread attention because of their effectiveness in solving few-shot classification problems. However, the scarcity of data frequently results in suboptimal embeddings, causing a discrepancy between anticipated class prototypes and those derived from the support set. These problems severely limit the generalizability of such methods, necessitating further development of Few-Shot Learning (FSL). In this study, we propose the Contrastive Prototype Network (CPN) consisting of three components: (1) Contrastive learning proposed as an auxiliary path to reduce the distance between homogeneous samples and amplify the differences between heterogeneous samples, thereby enhancing the effectiveness and quality of embeddings; (2) A pseudo-prototype strategy proposed to address the bias in prototypes, whereby the pseudo prototypes generated using query set samples are integrated with the initial prototypes to obtain more representative prototypes; (3) A new data augmentation technique, mixupPatch, introduced to alleviate the issue of insufficient data samples, whereby enhanced images are generated by blending the images and labels from different samples, to increase the number of samples. Extensive experiments and ablation studies conducted on five datasets demonstrated that CPN achieves robust results against recent solutions.

Effect of Graphene Embedment on Fiber–Matrix Interface and Tensile Properties of FRCM Composites

Effect of Graphene Embedment on Fiber–Matrix Interface and Tensile Properties of FRCM Composites

Find the “best evidence” of the relationship between network embeddedness and enterprise disruptive innovation performance

PurposeIn the context of open innovation, more and more enterprises are leveraging innovation networks to drive disruptive innovation performance, but there is no consensus on the relationship between network embeddedness and enterprise disruptive innovation performance. This paper aims to systematically explore the relationship between them.Design/methodology/approachThis paper constructs a multi-level network embeddedness model and uses 58 independent studies as samples to explore the relationship between multi-level network embeddedness and enterprise disruptive innovation performance by meta-analysis.FindingsFirst, network embeddedness at the enterprise and regional levels will promote the improvement of disruptive innovation performance. Although industrial relationship embeddedness will promote the improvement of disruptive innovation performance, its structural embeddedness will bring negative effects. Second, in terms of mediating effect, policy-oriented support will promote the relationship between network embeddedness and disruptive innovation performance at the enterprise and industry levels. Compared with large enterprises, small- and medium-sized enterprises will have more advantages in the performance of multi-level network embedding and disruptive innovation performance. Under the subjective performance measurement method, the promotion effect of multi-level network embedding is more prominent.Research limitations/implicationsThis study enriches the theoretical research of network embeddedness and disruptive innovation and provides management enlightenment for the network embeddedness strategy of enterprise disruptive innovation. Limited by data samples and article length, future research can further expand literature samples to test the stability of variable relationships and test the moderating effects of more internal and external factors.Originality/valueFirst, it constructs a theoretical analysis model of “point-line-surface” multi-level network embedding and disruptive innovation performance of enterprises and expands the theoretical analysis framework of network embedding and disruptive innovation performance. The second is to explore the influence mechanism of multi-level network embeddedness and enterprise disruptive innovation performance. Third, it deepens the theoretical understanding of the moderating variables of the impact of network embeddedness and enterprise disruptive innovation performance.

Effectiveness of ELMo embeddings, and semantic models in predicting review helpfulness

Online product reviews (OPR) are a commonly used medium for consumers to communicate their experiences with products during online shopping. Previous studies have investigated the helpfulness of OPRs using frequency-based, linguistic, meta-data, readability, and reviewer attributes. In this study, we explored the impact of robust contextual word embeddings, topic, and language models in predicting the helpfulness of OPRs. In addition, the wrapper-based feature selection technique is employed to select effective subsets from each type of features. Five feature generation techniques including word2vec, FastText, Global Vectors for Word Representation (GloVe), Latent Dirichlet Allocation (LDA), and Embeddings from Language Models (ELMo), were employed. The proposed framework is evaluated on two Amazon datasets (Video games and Health & personal care). The results showed that the ELMo model outperformed the six standard baselines, including the fine-tuned Bidirectional Encoder Representations from Transformers (BERT) model. In addition, ELMo achieved Mean Square Error (MSE) of 0.0887 and 0.0786 respectively on two datasets and MSE of 0.0791 and 0.0708 with the wrapper method. This results in the reduction of 1.43% and 1.63% in MSE as compared to the fine-tuned BERT model on respective datasets. However, the LDA model has a comparable performance with the fine-tuned BERT model but outperforms the other five baselines. The proposed framework demonstrated good generalization abilities by uncovering important factors of product reviews and can be evaluated on other voting platforms.

The Impact of Global Value Chain Embedment on Energy Conservation and Emissions Reduction:Theory and Empirical Evidence

Abstract An in-depth investigation into the effect of embedment in global value chain (GVC) on energy conservation and emissions reduction is of great significance for scientifically assessing the environmental impact of GVC participation, and promoting high-quality development in China. This paper incorporates GVC embedment, energy consumption and carbon emissions into the same analysis framework for the first time. Based on the WIOD database, this paper theoretically and empirically examines the impact and mechanism of global value chain embeddedness on carbon emission reduction from two dimensions: energy consumption intensity and energy consumption structure. The study found that GVC embedment significantly reduced the industry’s carbon emission intensity; developing economies’ embedment in GVC helped reduce their carbon emission intensity, while the effect was not obvious in developed economies. GVC embedment had a significant inhibitory effect on the carbon emissions in both upstream and downstream industries, but not conducive to carbon reduction of low-tech manufacturing. The mechanism test shows that the GVC embedment not only exhibits the dual effects of energy conservation and emissions reduction, but also has a significant impact on carbon emissions by reducing the energy consumption intensity and improving the energy consumption structure.