Search

Recommender systems are widely used in e-commerce, movies, music, social media, and other fields because of their personalized recommendation functions. The recommendation algorithm is used to capture user preferences, item characteristics, and the items that users are interested in are recommended to users. Matrix factorization is widely used in collaborative filtering algorithms because of its simplicity and efficiency. However, the simple dot-product method cannot establish a nonlinear relationship between user latent features and item latent features or make full use of their personalized information. The model of a neural network combined with an attention mechanism can effectively establish a nonlinear relationship between the potential features of users and items and improve the recommendation accuracy of the model. However, it is difficult for the general attention mechanism algorithm to solve the problem of attention interaction when the number of features between the users and items is not the same. To solve the above problems, this paper proposes an attention interaction matrix factorization (AIMF) model. The AIMF model adopts a symmetric structure using MLP calculation. This structure can simultaneously extract the nonlinear features of user latent features and item latent features, thus reducing the computation time of the model. In addition, an improved attention algorithm named slide-attention is included in the model. The algorithm uses the sliding query method to obtain the user's attention to the latent features of the item and solves the interaction problem among different dimensions of the user, and the latent features of the item. [ABSTRACT FROM AUTHOR]

To improve the energy conversion and efficiency of hydrogen-fueled combustion, a micro recuperative planar with porous media is proposed. Numerical simulation model of premixed H 2 /air combustion is verified by experimental test, then orthogonal analysis of multi-factors, i.e., porous media setting and flow rate, on the combustion characteristics and heat transmission are compared and analyzed. The results indicate that combustion characteristics and heat release are strongly impacted by porous media, that is, preheating of unburnt reactants and heat transmission of gases and gas-solid are improved via the heat recirculation in/around porous media. For instance, the difference of mean wall temperatures of the combustor with/without porous media reaches 4.5% at m f = 6.67 × 10−7 kg/s. So, the inserted porous media boosts the burner radiation temperature, and porous media dimensions and porosity affect the thermal performance of the burner, reducing exhaust gas temperature and improving energy efficiency. Besides, orthogonal analysis is also employed to understand the coupling effects of multi-factors on burning and heat transmission, the effect degree on T m is ranked by m f , L , H and φ. Hence, an optimal burning setting is obtained, i.e., combustor C4.8–6.8, where T m is increased by 4.6% compared to the combustor C3-3 at m f = 9.17 × 10−7 kg/s. It achieves the maximum radiation efficiency 35.26% at m f = 6.34 × 10−7 kg/s. • A micro combustor with porous media is proposed for thermophotovoltaic. • Effects of porous media setting and flow rates on combustion are investigated. • A higher radiation performance can be obtained in the burner with porosity 0.7. • Comprehensive analysis and grey correlation of multi-factors are conducted. • Optimal burner setting is achieved with the maximum radiation efficiency 35.26%. [ABSTRACT FROM AUTHOR]

Condensed films of functional luminophores dominated by the magnitude and dimensionality of the intermolecular interactions play important roles in sensing performance. However, controlling the molecular assembly and regulating photophysical properties remain challenging. In this study, a new luminophore, ortho‐PBI‐Au, was synthesized by anchoring a cyclometalated alkynyl‐gold(III) unit at the ortho‐position of perylene bisimide. An unprecedented T‐type packing model driven by weak Au‐π interaction and Au−H bonds was observed, laying foundation for striking properties of the luminophore. Controlled assembly of ortho‐PBI‐Au at the air‐water interface, realized using the classical Langmuir–Schaeffer technique, afforded the obtained luminescent films with different packing structures. With an optimized film, sensitive, selective, and rapid detection of a hazardous new psychoactive substance, phenylethylamine (PEA), was achieved. The detection limit, response time, and recovery time were <4 ppb, <1 s, and <5 s, respectively, surpassing the performance of the PEA sensors known thus far. The relationship between the characters of films and the sensing performance was systematically examined by grey relational analysis (GRA). The present study suggests that designing novel molecular aggregation with definite adlayer structure is a crucial strategy to enhance the sensing performance, which could be favorable for the film‐based fluorescent sensors. [ABSTRACT FROM AUTHOR]

The behavior of a nanosecond pulsed sliding discharge plasma actuator with the ambient pressure from 27 to 101 kPa is experimentally investigated. The electric characteristics, discharge morphologies, and surface temperature distribution of the actuator supplied by constant voltages are studied under different pressure conditions. The threshold pressure for sliding discharge establishment is 54–75 kPa in this paper. The results show that the positive peak value of the current tends to increase as the pressure decreases. Time-integrated discharge images indicate that plasma luminosity and uniformity are enhanced under low-pressure conditions. The discharge morphology of the actuator at 27 kPa displays an alternating distribution of bright and dark stripes. The infrared thermal results demonstrate that decreasing the pressure increases the actuator's surface temperature, intensifies the intensity of spanwise temperature oscillations, and improves the homogeneity in the streamwise direction. [ABSTRACT FROM AUTHOR]