A Rating Prediction Recommendation Model Combined with the Optimizing Allocation for Information Granularity of Attributes

In recent years, graph neural networks (GNNS) have been demonstrated to be a powerful way to learn graph data. The existing recommender systems based on the implicit factor models mainly use the interactive information between users and items for training and learning. A user–item graph, a user–attribute graph, and an item–attribute graph are constructed according to the interactions between users and items. The latent factors of users and items can be learned in these graph structure data. There are many methods for learning the latent factors of users and items. Still, they do not fully consider the influence of node attribute information on the representation of the latent factors of users and items. We propose a rating prediction recommendation model, short for LNNSR, utilizing the level of information granularity allocated on each attribute by developing a granular neural network. The different granularity distribution proportion weights of each attribute can be learned in the granular neural network. The learned granularity allocation proportion weights are integrated into the latent factor representation of users and items. Thus, we can capture user-embedding representations and item-embedding representations more accurately, and it can also provide a reasonable explanation for the recommendation results. Finally, we concatenate the user latent factor-embedding and the item latent factor-embedding and then feed it into a multi-layer perceptron for rating prediction. Extensive experiments on two real-world datasets demonstrate the effectiveness of the proposed framework.

STUDY ON GRANULARITY DISTRIBUTION OF POWDER BY FRACTAL MODELS

The granularity distribution of powder is shown to have the fractal characters. Based on the fractal theory, three algebraic expressions for the fractal dimension, the cumulative probability and the mass distribution of particles were obtained. Three expressions are the functions of the concentration of particles, the diameter of particles, the maximum and the minimum particle diameters. No additional empirical constant is introduced in these expressions. These fractal models contain less empirical constants than the conventional correlations. The fractal behavior about coal particles was discussed in detail. Simulations were also performed by fractal Monte Carlo technique. The proposed technique may provide us a new approach to analyze the granularity distribution of powder.

Effect of Bimodal Granularity Distribution on the Properties of Silica-Based Ceramic Cores

In this work, the effects of the bimodal particle size distribution of fused silica on the properties of silica-based ceramic cores have been investigated. In order to simulate a casting process condition, the core specimens were sintered at 1220°C, tested at above 1500°C. Three point bending tests were carried out on all the prepared specimens. Phase evolution and microstructure were investigated by XRD and SEM respectively. The results showed that with the bimodal granularity distribution broaden and the content of coarse particles increasing, the contraction, flexural strength and creep deformation of ceramic cores decreased linearly, as the content of mineralizer was determined. Coarse particles served as the skeleton to relieve the contraction in ceramic matrix and avoided sharp decrease of flexural strength. The high temperature properties were largely determined by the combined effects of crystallization and skeleton network. The crystallization depended mostly on the fine particles, and the skeleton network was prerequisite to avoid creep deformation.

The Study of the Appropriate Reclamation Method for the Complete Reclamation of the Used Clay Sand

The paper studies the process adaptability of the complete reclamation for the used clay sand by countercurrent rotor reclamation, grinding wheel reclamation, air flow reclamation, and vibration reclamation, and found the complete reclamation process of the used clay sand by high temperature baking, reclamation and micro powder separation. Compared with the base sand, the completely reclaimed clay sand obtained by using countercurrent rotor reclaimer, grinding wheel reclaimer and air flow reclaimer has no change of the granularity distribution, lower mud content and higher acid demand value. And the use of vibration reclaimer can only obtain the incompletely reclaimed clay sand that has no change of granularity distribution and high mud content and acid demand value compared with the base sand.

Study on Molecule Dimension of Different Granularity Konjac Glucomannan

In this paper, the relativity on the granularity with different mesh of konjac glucomannan (KGM) and molecular dimension was discussed. Laser particle size analyzer, laser light scattering and digital polarimeter were used to analyze the granularity distribution, molecular characteristics and polarity of KGM. The results showed that the granularity of KGM after gradient refining was gradually decreased and KGM was well dispersed in ethanol(D50=13.11μm).The Mw of KGM was gradually decreased with its granularity decrease and its molecular conformation was from globular to linear structure. The space asymmetry of KGM at 80-100 meshes was minimum and its polarity was maximum. The primary establishment on the relativity offered the references to KGM theoretical research and practical application.

Treatment of Oily Sludge from Floatation Process: Floc Breakage and Dewatering

The aim is to treat oily sludge from floatation process of oil produced water by deflocculation. Based upon analyzing the granularity distribution and dewater efficiency of the flocs, floc breakage and dewater of oily sludge was carried out. The results indicated that the coagulant skeleton builder in sludge could be broken by changing the coalescent state, and deflocculation could disrupt skeleton of the floc and release interstitial water in sludge. The highest dewatering efficiency is about 81.2% at the pH of 4.5 by gravity sedimentation. The sludge had a broader range of granularity distribution and a smaller volume-average diameter. Moreover, the median diameter d (0.5) of the treated sludge was 36.0 μm, which was smaller than 73.1 μm of raw sample. Meanwhile, capillary suction time (CST) significantly decreased with pH adjustment. A wastewater removal efficiency of 90.99% for the treated sludge by pressure filtration could be achieved. The method would be a potential technique to be applied to other kinds of floatation sludge.