Nutrition-Related Knowledge Graph Neural Network for Food Recommendation.

Food recommendation systems are becoming increasingly vital in modern society, given the fast-paced lifestyle and diverse dietary habits. Existing research and implemented solutions often rely on user preferences and past behaviors for recommendations, which poses significant issues. Firstly, this approach inadequately considers the nutritional content of foods, potentially leading to recommendations that are overly homogeneous and lacking in diversity. Secondly, it may result in repetitive suggestions of the same types of foods, thereby encouraging users to develop unhealthy dietary habits that could adversely affect their overall health. To address this issue, we introduce a novel nutrition-related knowledge graph (NRKG) method based on graph convolutional networks (GCNs). This method not only enhances users' ability to select appropriate foods but also encourages the development of healthy eating habits, thereby contributing to overall public health. The NRKG method comprises two key components: user nutrition-related food preferences and recipe nutrition components. The first component gathers nutritional information from recipes that users show interest in and synthesizes these data for user reference. The second component connects recipes with similar nutritional profiles, forming a complex heterogeneous graph structure. By learning from this graph, the NRKG method integrates user preferences with nutritional data, resulting in more accurate and personalized food recommendations. We evaluated the NRKG method against six baseline methods using real-world food datasets. In the 100% dataset, the five metrics exceeded the performance of the best baseline method by 2.8%, 5.9%, 1.5%, 9.7%, and 6.0%, respectively. The results indicate that our NRKG method significantly outperforms the baseline methods, including FeaStNet, DeepGCN, GraphSAGE, GAT, UniMP, and GATv2, demonstrating its superiority and effectiveness in promoting healthier and more diverse eating habits. Unlike these baseline methods, which primarily focus on hierarchical information propagation, our NRKG method offers a more comprehensive approach by integrating the nutritional information of recipes with user preferences.

Traffic prediction in SDN for explainable QoS using deep learning approach.

The radical increase of multimedia applications such as voice over Internet protocol (VOIP), image processing, and video-based applications require better quality of service (QoS). Therefore, traffic Predicting and explaining the prediction models is essential. However, elephant flows from those applications still needs to be improved to satisfy Internet users. Elephant flows lead to network congestion, resulting in packet loss, delay and inadequate QoS delivery. Recently, deep learning models become a good alternative for real-time traffic management. This research aims to design a traffic predicting model that can identify elephant flows to prevent network congestion in advance. Thus, we are motivated to develop elephant flow prediction models and explain those models explicitly for network administrators' use in the SDN network. H2O, Deep Autoencoder, and autoML predicting algorithms, including XGBoost, GBM and GDF, were employed to develop the proposed model. The performance of Elephant flow prediction models scored 99.97%, 99.99%, and 100% in validation accuracy of under construction error of 0.0003952, 0.001697, and 0.00000408 using XGBoost, GBM, and GDF algorithms respectively. The models were also explicitly explained using Explainable Artificial Intelligence. Accordingly, packet size and byte size attributes need much attention to detect elephant flows.

Recovery of copper from electroplating sludge using integrated bipolar membrane electrodialysis and electrodeposition.

Electroplating sludge, though a hazardous waste, is a valuable resource as it contains a large amount of precious metals. In this study, copper was recovered from the electroplating sludge using a technology that integrates bipolar membrane electrodialysis (BMED) and electrodeposition. The experimental results showed that Cu2+ in the electroplating sludge was successfully separated and concentrated in the BMED system without adding any chemical reagents; the concentrated Cu2+ was recovered in the form of copper foil in an electrodeposition system. Current density clearly affected the Cu2+ separation and concentration in the BMED system; the current density, solution pH and Cu2+ concentration drastically affected the Cu2+ electrodeposition ratio and the morphology and purity of the obtained copper foil. Under the optimised experimental conditions, 96.4% of Cu2+ was removed from the electroplating sludge and 65.4% of Cu2+ was recovered in the foil form. On increasing the number of electroplating sludge compartments from one to two and three, the current efficiency for recovering Cu2+ increased from 17.4% to 28.5% and 35.2%, respectively, and the specific energy consumption decreased from 11.3 to 6.7 and 5.3 kW h/kg of copper, respectively. The purity of the copper foil was higher than 99.5%. Thus, the integrated technology can be regarded as an effective method for recovering copper from electroplating sludge.

Nickel recovery from electroplating sludge via bipolar membrane electrodialysis.

In this study, nickel (Ni) was recovered from electroplating sludge in the form of Ni(OH)2 using a bipolar membrane electrodialysis (BMED) system. The results showed that the H+ generated by the bipolar membrane could effectively desorb Ni from the sludge to the solution and the solution pH considerably affected Ni desorption. The desorption process can be described using the first-order kinetic model. The current density and solid/liquid ratio (m/v) considerably affected Ni recovery. Moreover, 100% of Ni was removed from the electroplating sludge and 93.5% of Ni was recovered after 28 h under a current density of 20 mA/cm2, a solid/liquid ratio of 1.0:15 and an electroplating-sludge particle size of 100 mesh. As the number of electroplating compartments increased from one to two and three, the current efficiency for recovering Ni changed from 12.1% to 11.8% and 11.9%, respectively, and the specific energy consumption decreased from 0.064 to 0.048 and 0.039 kW·h/g, respectively. Fourier-transform infrared spectroscopy and Raman spectroscopy showed that the precipitate obtained in this study is similar to commercial Ni(OH)2 and the purity of Ni(OH)2 in the obtained precipitate was 79%. Thus, the results showed that the BMED system is effective for recovering Ni from electroplating sludge.

Securing federated learning with blockchain: a systematic literature review.

Federated learning (FL) is a promising framework for distributed machine learning that trains models without sharing local data while protecting privacy. FL exploits the concept of collaborative learning and builds privacy-preserving models. Nevertheless, the integral features of FL are fraught with problems, such as the disclosure of private information, the unreliability of uploading model parameters to the server, the communication cost, etc. Blockchain, as a decentralized technology, is able to improve the performance of FL without requiring a centralized server and also solves the above problems. In this paper, a systematic literature review on the integration of Blockchain in federated learning was considered with the analysis of the existing FL problems that can be compensated. Through carefully screening, most relevant studies are included and research questions cover the potential security and privacy attacks in traditional federated learning that can be solved by blockchain as well as the characteristics of Blockchain-based FL. In addition, the latest Blockchain-based approaches to federated learning have been studied in-depth in terms of security and privacy, records and rewards, and verification and accountability. Furthermore, open issues related to the combination of Blockchain and FL are discussed. Finally, future research directions for the robust development of Blockchain-based FL systems are proposed.

Recovery of nickel in the form of Ni(OH)2 from plating wastewater containing Ni-EDTA using bipolar membrane electrodialysis.

Ni is often present in plating wastewater as a complexing state. It is difficult to remove this Ni using traditional chemical precipitation technology. In this study, a bipolar membrane electrodialysis system was used to recover Ni in the form of Ni(OH)2 from plating wastewater containing Ni-ethylenediaminetetraacetic acid (Ni-EDTA) without adding chemical reagents. The stable structure of Ni-EDTA can be destroyed by H+ produced by the bipolar membrane to obtain free Ni2+, which can combine with OH- produced by the bipolar membrane to form Ni(OH)2. When the electrolyte Na2SO4 concentration, current density and initial Ni-EDTA concentration were 0.2 mol/L, 16 mA/cm2 and 1000 mg/L, respectively, 99.0% of Ni-EDTA was removed after 32 h. When the system was used to treat actual plating wastewater, 92.1% of Ni-EDTA was removed and 88.7% was recovered. When the number of wastewater compartments in the system was increased from one to three, the current efficiency increased from 1.7% to 5.8%, and the specific energy consumption decreased from 0.39 to 0.19 kW h/g. The results of an X-ray diffraction study indicate that the Ni(OH)2 obtained in this study is similar to commercial Ni(OH)2. Moreover, the recovery mechanism of Ni-EDTA was analysed. Thus, bipolar membrane electrodialysis can be regarded as an effective method to recover Ni from wastewater containing Ni-EDTA.

Functional analysis revealed the involvement of ZmABCB15 in resistance to rice black-streaked dwarf virus infection.

BACKGROUND: Maize rough dwarf disease (MRDD), caused by rice black-streaked dwarf virus (RBSDV) belonging to the Fijivirus genus, seriously threatens maize production worldwide. Three susceptible varieties (Ye478, Zheng 58, and Zhengdan 958) and two resistant varieties (P138 and Chang7-2) were used in our study. RESULTS: A set of ATP-binding cassette subfamily B (ABCB) transporter genes were screened to evaluate their possible involvements in RBSDV resistance. In the present study, ZmABCB15, an ABCB transporter family member, was cloned and functionally identified. Expression analysis showed that ZmABCB15 was significantly induced in the resistant varieties, not in the susceptible varieties, suggesting its involvement in resistance to the RBSDV infection. ZmABCB15 gene encodes a putative polar auxin transporter containing two trans-membrane domains and two P-loop nucleotide-binding domains. Transient expression analysis indicated that ZmABCB15 is a cell membrance localized protein. Over-expression of ZmABCB15 enhanced the resistance by repressing the RBSDV replication ratio. ZmABCB15 might participate in the RBSDV resistance by affecting the homeostasis of active and inactive auxins in RBSDV infected seedlings. CONCLUSIONS: Polar auxin transport might participate in the RBSDV resistance by affecting the distribution of endogenous auxin among tissues. Our data showed the involvement of polar auxin transport in RBSDV resistance and provided novel mechanism underlying the auxin-mediated disease control technology.

Recovery of trivalent and hexavalent chromium from chromium slag using a bipolar membrane system combined with oxidation.

Chromium slag (CS) with large quantities of multivalent Cr species (III and VI) generated during chromium salt production is hazardous to nature and living organisms. Furthermore, CS discharge leads to considerable resource wastage. Herein, a bipolar membrane electrodialysis (BMED) system was employed along with hydrogen peroxide (H2O2) oxidation for simultaneously recovering Cr(III) and Cr(VI) from CS in the form of Na2CrO4. A bipolar membrane was used to produce OH- under a direct electric field, providing an alkaline environment for the oxidative conversion of Cr(III) to Cr(VI) in the presence of H2O2, followed by the recovery of Cr(III) and Cr(VI) as Na2CrO4. The effect of H2O2content on Cr(III) oxidation and that of the current density on chromium recovery, current efficiency and specific energy consumption were investigated. Moreover, the morphology of chromium in CS before and after the BMED treatment was analysed. The H2O2 content affected the Cr oxidation rate from Cr(III) to Cr(VI). The current density affected chromium removal, current efficiency and specific energy consumption. At a current density of 2 mA/cm2, the total chromium recovery exceeded 67% and the remaining chromium was mainly in the residual state (RES). When the number of CS compartments increased, the current efficiency was enhanced and the specific energy consumption decreased. Binding state analysis show that Cr(III) and different species of Cr(VI) could be transformed into exchangeable Cr(VI) after H2O2 oxidation and BMED treatment. After the treatment, 92% of the remaining chromium in CS was in the RES. Thus, the employed method can effectively recover chromium from CS and other chromium-contaminated solid waste.

Recovery of phosphate and ammonium nitrogen as struvite from aqueous solutions using a magnesium-air cell system.

The addition of alkaline and magnesium sources during the recovery of NH4+ and PO43- in the form of struvite using the traditional struvite precipitation method increases the production cost. To solve this problem, a magnesium-air cell (MAC) system was used herein to recover NH4+ and PO43- as struvite from wastewater using a magnesium strip (Mg2+) and the oxygen adsorbed on the surface of a titanium plate (OH-) as the anode and cathode, respectively. Experimental parameters (i.e. initial solution pH, temperature, NH4+/PO43- molar ratio, NH4+ and PO43- initial concentrations and stirring intensity) were found to affect the removal rate of NH4+ and PO43-. The presence of Ca2+ decreased the struvite purity. At Ca2+/PO43- ratios of 0:1 and 0.5:1, the purity of the obtained struvite after 6 h was 93.8% and 58.9%, respectively. Struvite with a purity of 95.7%, electricity with an average output power of 2.53 mW, and an energy density of 1.05 W/m2 were obtained when the proposed system was used to recover NH4+ and PO43- from an actual supernatant of domestic sludge anaerobic digestion. Scanning electron microscopy-energy-dispersive X-ray spectroscopy, X-ray diffraction, Fourier transform infrared spectroscopy, Raman spectroscopy and thermogravimetric analyses showed that the obtained struvite exhibited almost the same physicochemical properties as commercial struvite. Thus, the MAC system can be regarded as an effective method for recovering NH4+ and PO43- in the form of struvite from wastewater.

[CT and MRI imaging features of peripheral primitive neuroectodermal tumors].

OBJECTIVE: To investigate the CT and MRI features of peripheral primitive neuroectodermal tumors (pPNETs) and evaluate its diagnostic value. METHODS: The clinicopathological data of 9 surgically treated patients with peripheral primitive neuroectodermal tumors confirmed by pathology were collected, spiral CT (4/9) and MRI (6/9) plain scanning and dynamic enhancement scanning were performed preoperatively. Both CT and MRI scannings were performed in 1 case. Those CT and MR images were retrospectively reviewed and analyzed together with clinicopathological findings. RESULTS: The 9 lesions were located in skeletal muscles (n = 6), pelvic cavity (n = 2) and thoracic cavity (n = 1). The tumor size was 7.4-18.3 cm in diameter with a mean diameter of 11.6 cm. The shape of those lesions was round or ellipse (4 lesions) and irregular (5 lesions). The tumor usually presented as ill-defined masses, with homogeneous (n = 2) or inhomogeneous density (n = 7). Seven cases, including the 3 lesions located in the chest and pelvis, showed obvious necrosis and multilocular cyst formation. The tumors showed iso-density as that of the adjacent muscles on CT plain scans and moderate heterogeneous enhancement after intravenous injection of contrast agents. The features of the tumors on the MRI including slightly low signal intensity on SE T1-weighted imaging, iso-signal intensity or slightly high signal intensity on FSE T2-weighted imaging and heterogeneous dynamic delayed contrast-enhancement with obvious necrosis in most of them. Six cases had a lesion in the skeletal muscles, presented as a giant ill-defined masse surrounding bone and extended along neural route with bone destruction to varying degrees. CONCLUSION: Primitive neuroectodermal tumor is a kind of malignant tumor with proliferation of small, undifferentiated neuroectodermal cells, usually occurring in children or adolescent and frequently located in the extremities, chest cavity, pelvic cavity and chest wall. It typically presents as a large, ill-defined masse extending along neural route with heterogeneous and obvious enhancement after intravenous injection of contrast agents. The tumors located in the chest and pelvic cavities and some in the extremities show obvious necrosis and multilocular cyst formation, while some of the tumors in the extremities appear as homogeneous solid masses.

[Value of spiral CT in diagnosis of cystic renal cell carcinoma].

OBJECTIVE: To investigate the image features and the diagnostic value of spiral CT for cystic renal cell carcinoma. METHODS: The clinical data and CT manifestations of 17 operated and pathologically proven cystic renal cell carcinoma were retrospectively analyzed. There were 12 males and 5 females with an average age of 47.3 years (33 - 82 years). Plain and contrast CT scan (Siemens somatom) single layer sensation 16 layer spiral CT had been performed before operation. The image of artery phase (30 - 40 s), venous (60 - 70 s) and excretory (120 - 180 s) were respectively obtained after contrast administration. Various image reconstructions were done using Siemens Wizard workstation based on the raw images. RESULTS: It was found that 5 cystic renal cell cancers located in the right kidney and 12 in the left kidney. The long dimension of the tumor arranged from 21 - 100 mm with an average of 57 mm. The tumor looked like a round or round-like shape with density similar to fluid on plain CT scan. Some cystic renal carcinomas had a thick wall. Some had single or multiple cystic spaces filled with fluid of different densities. Some had infiltrated out of kidney surface or into renal sinus. Some showed enhanced nodules on the wall. CONCLUSION: Cystic renal cell carcinoma has its own specific morphologic features in spiral CT scan. Spiral CT may be very helpful in the diagnosis of cystic renal cell carcinoma before operation.

[Cloning and expression of ethylene receptor gene Ad-ETR1 from kiwifruit].

According to the conserved amino acid sequence from ethylene receptors in other plants, a pair of degenerate primers was designed and a 657-bp cDNA fragment encoding an ethylene receptor fruit was obtained by RT-PCR from ripening kiwifruit (Actinidia deliciosa cv. Bruno) (Fig. 1). The cDNA fragment encoding 219 amino acids was named Ad-ETR1, and its sequences shared high similarity at both nucleotide and polypeptide level with the sequences from the plants of Arabidopsis, tomato, persimmon, avocado, citrus and peach (Fig. 2, Table 1). Northern blot analysis indicated that the levels of Ad-ETR1 mRNA increased during kiwifruit ripening and reached peak at 144 h after treatment, then dropped immediately. The expression of Ad-ETR1 could be induced by ethylene treatment, while acetylsalicylic acid (ASA) treatment inhibited its expression (Fig. 4). In consistency with the changes of Ad-ETR1 mRNA, ethylene or ASA treatment had marked effects on the kiwifruit ethylene production and fruit softening (Fig. 3). The significance of these results was discussed.