The increasing risk of ground pressure disasters resulting from deep well mining highlights the urgent need for advanced monitoring and early warning systems.Ground pressure monitoring,supported by microseismic techno...The increasing risk of ground pressure disasters resulting from deep well mining highlights the urgent need for advanced monitoring and early warning systems.Ground pressure monitoring,supported by microseismic technology,plays a pivotal role in ensuring mine safety by enabling real-time identifi cation and accurate classification of vibration signals such as microseismic signals,blasting signals,and noise.These classifications are critical for improving the efficacy of ground pressure monitoring systems,conducting stability analyses of deep rock masses,and implementing timely and precise roadway support measures.Such eff orts are essential for mitigating ground pressure disasters and ensuring safe mining operations.This study proposes an artificial intelligence-based automatic classification network model for mine vibration signals.Based on conventional convolutional neural networks,the proposed model further incorporates long short-term memory(LSTM)networks and attention mechanisms.The LSTM component eff ectively captures temporal correlations in time-series mining vibration data,while the attention mechanism enhances the models’ability to focus on critical features within the data.To validate the eff ectiveness of our proposed model,a dataset comprising 480,526 waveform records collected in 2022 by the microseismic monitoring system at Guangxi Shanhu Tungsten Mine was used for training,validation,and testing purposes.Results demonstrate that the proposed artifi cial intelligence-based classifi cation method achieves a higher recognition accuracy of 92.21%,significantly outperforming traditional manual classification methods.The proposed model represents a signifi cant advancement in ground pressure monitoring and disaster mitigation.展开更多
The rapid advancement of modern science and technology,coupled with the recent surge in new-energy electric vehicles,has significantly boosted the demand for lithium.This has promoted the development and efficient uti...The rapid advancement of modern science and technology,coupled with the recent surge in new-energy electric vehicles,has significantly boosted the demand for lithium.This has promoted the development and efficient utilization of lepidolite as a lithium source.Therefore,the processes for the flotation of lepidolite have been studied in depth,particularly the development and use of lepidolite flotation collectors and the action mechanism of the collectors on the lepidolite surface.Based on the crystal-structure characteristics of lepidolite minerals,this review focuses on the application of anionic collectors,amine cationic collectors(primary amines,quaternary ammonium salts,ether amines,and Gemini amines),and combined collectors to the flotation behavior of lepidolite as well as the adsorption mechanisms.New directions and technologies for the controllable flotation of lepidolite are proposed,including process improvement,reagent synthesis,and mechanistic research.This analysis demonstrates the need for the further study of the complex environment inside lepidolite and pulp.By using modern analytical detection methods and quantum chemical calculations,research on reagents for the flotation of lepidolite has expanded,providing new concepts and references for the efficient flotation recovery and utilization of lepidolite.展开更多
NiFe-layered double hydroxides(NiFe-LDHs)are among the most promising earth-abundant electrocatalysts for the oxygen evolution reaction(OER)in alkaline media.However,their practical application is hindered by intrinsi...NiFe-layered double hydroxides(NiFe-LDHs)are among the most promising earth-abundant electrocatalysts for the oxygen evolution reaction(OER)in alkaline media.However,their practical application is hindered by intrinsic activity limitations and poor stability,primarily due to the asymmetric adsorption of oxygen intermediates.To overcome this,the binding strength must be synergistically tuned to a moderate level to optimize catalytic performance.Here,we engineered NiFeCoCr LDH through Co doping to enhance electrical conductivity and controlled Cr leaching to introduce cationic vacancies for modulating intermediate binding strength in NiFe LDH.X-ray absorption near-edge structure and extended X-ray absorption fine structure analyses reveal that NiFe-LDH with Co doping and Cr vacancies modulates the Ni oxidation state and local coordination environment,leading to a balanced electronic structure and enhanced structural complexity around the Ni sites.Additionally,these vacancies can trap OH^(-)/H_(2)O species,which can serve as a reservoir for OH^(-) transfer,facilitating the rapid formation of OER intermediates and enhancing catalytic performance at high current densities.As a result,V_(Cr)-NiFeCo LDH achieves 1.6 A cm^(-2)current density at 1.7 V vs.RHE while maintaining stable operation for over 1000 h at 500 mA cm^(-2).Density functional theory(DFT)calculations validate the synergistic effects of Co doping and Cr-induced vacancies on intermediate binding energies and improved OER kinetics.Overall,this work presents a rational design strategy to simultaneously enhance the activity and durability of NiFe-based OER catalysts for their application in high-performance alkaline water electrolysis.展开更多
The development of synthetic hybrid biological systems integrating photosynthetic organisms with organic-abiotic functional materials holds significant promise for enhancing photosynthetic processes.The artificial reg...The development of synthetic hybrid biological systems integrating photosynthetic organisms with organic-abiotic functional materials holds significant promise for enhancing photosynthetic processes.The artificial regulation of the state transition between photosystem I(PSI)and photosystem II(PSII)represents a strategic and promising approach for improving the efficiency of natural photosynthesis.In this study,we demonstrate that poly(benzimidazolium-phenylthiophene)(CP4)featuring a flexible cationic backbone exhibits superior ultraviolet light-harvesting capability.The polymer CP4 enhanced PSI activity in Chlorella pyrenoidosa(C.pyrenoidosa),subsequently promoting PSII activity and augmenting overall photosynthetic performance.During light-dependent reactions,CP4 significantly accelerated photosynthetic electron transfer,resulting in a 330%increase in the production of oxygen and 93%and 96%increases in the ATP and NADPH contents,respectively.In the context of dark reactions,CP4 facilitated the conversion and utilization of light energy,leading to a 6%increase in both carbohydrate and protein contents.These findings indicate that synthetic light-harvesting polymer materials exhibit considerable application potential in the field of biomass production through enhancement of natural photosynthetic efficiency.展开更多
Cataract is the leading cause of reversible blindness worldwide,affecting millions,particularly the elderly.Over 65 million people suffer from significant visual impairment due to cataracts,with the burden being highe...Cataract is the leading cause of reversible blindness worldwide,affecting millions,particularly the elderly.Over 65 million people suffer from significant visual impairment due to cataracts,with the burden being highest in low-and middle-income countries where access to surgery is limited.Cataract surgery,one of the most commonly performed and cost-effective procedures,has evolved significantly.Traditional extracapsular cataract extraction(ECCE)has been largely replaced by phacoemulsifi cation,which uses ultrasonic energy through a small incision,reducing recovery time and complications.More recently,femtosecond laser-assisted cataract surgery(FLACS)has emerged,off ering enhanced precision but with ongoing evaluation of its cost-eff ectiveness.Intraocular lenses(IOLs)now allow for customized visual outcomes,addressing distance,near,and intermediate vision.Despite its safety,cataract surgery can still result in complications such as corneal edema and posterior capsular opacifi cation,requiring careful surgical management and patient education.展开更多
Sodium layered oxides stand out as one of the most promising cathodes for sodium-ion batteries due to their high energy density,elemental abundance,and scalability.However,their practical applications are restricted b...Sodium layered oxides stand out as one of the most promising cathodes for sodium-ion batteries due to their high energy density,elemental abundance,and scalability.However,their practical applications are restricted by interplanar gliding,cation migration,and the formation of intragranular microcracks,which collectively lead to rapid structural degradation and capacity loss.Herein,we rationally design an ultrastable O3-type Na_(0.94)Ca_(0.03)Ni_(1/3)Fe_(1/3)Mn_(1/3)O_(2) cathode,in which Ca^(2+)cations act as pillars within the NaO_(2)slabs,suppressing the irreversible phase transitions and Na/TM cation migration commonly observed in layered oxides.Multiscale in situ and ex situ techniques,combined with post-mortem analysis,reveal that the Ca-pillared pinning effect not only effectively suppresses the interplanar gliding and stress accumulation within the crystal phase but also restrains Na/TM cation migration and surface reconstruction in near-surface regions.Benefiting from the combined effects of structural stabilization,the Ca-pillared cathode exhibits a superior cycling stability,retaining 81.6%of its capacity after 1000 cycles at 2 C within the voltage range of 2.0-4.0 V,along with significantly enhanced wide-temperature(from-40 to 80℃)performance.This work highlights another critical role of Ca pillars in suppressing cation migration and surface structural degradation beyond preventing adverse interplanar gliding,offering valuable insights for designing long-life and wide-temperature layered oxide cathodes.展开更多
Converting CO_(2) into methanol(CH_(3)OH),a high-value-added liquid-phase product,through efficient and highly selective photocatalysis remains a significant challenge.Herein,we present a straightforward cation exchan...Converting CO_(2) into methanol(CH_(3)OH),a high-value-added liquid-phase product,through efficient and highly selective photocatalysis remains a significant challenge.Herein,we present a straightforward cation exchange strategy for the in-situ growth of BiVO_(4) on an InVO_(4) substrate to generate a Z-scheme heterojunction of InVO_(4)/BiVO_(4) .This in-situ partial transformation approach endows the InVO_(4)/BiVO_(4) heterojunction with a tightly connected interface,resulting in a significant improvement in charge separation efficiency between InVO_(4) and BiVO_(4).Moreover,the construction of the heterojunction reduces the formation energy barrier of the ^(*)COOH intermediate during the photoreduction of CO_(2) and increases the desorption energy barrier of the ^(*)CO intermediate,facilitating the deep reduction of ^(*)CO.Consequently,the InVO_(4)/BiVO_(4) heterojunction is capable of photocatalytic CO_(2) reduction to CH_(3)OH with high efficiency and selectivity.Under conditions where water serves as the electron source and a light intensity of 100 m W/cm^(2),the yield of CH_(3)OH reaches 130.5 μmol g^(-1)h^(-1) with a selectivity of 92 %,outperforming photocatalysts reported under similar conditions.展开更多
Background:Home accessibility modifi cations are crucial for promoting independent living and quality of life among persons with disabilities.While developed countries have established comprehensive policy frameworks,...Background:Home accessibility modifi cations are crucial for promoting independent living and quality of life among persons with disabilities.While developed countries have established comprehensive policy frameworks,developing nations like China face unique challenges in program design and implementation.Objective:This study conducts a systematic comparative analysis of home accessibility modification policies across China,Japan,Germany,and Sweden,identifying key policy dimensions and proposing evidence-based recommendations for strengthening China’s policy framework.Methods:We employed a multi-dimensional analytical framework examining legislative foundations,eligibility criteria,funding mechanisms,and service delivery models.Data were collected from primary legislation,governmental regulations,official statistics,and peer-reviewed literature.Results:Significant cross-national variations exist in policy approaches.Japan and Germany utilize social insurance models with standardized assessments,Sweden adopts a universal rights-based approach,while China employs a targeted assistance model focused on economically disadvantaged households.China completed 1.28 million household renovations during its 14th Five-Year Plan,demonstrating strong implementation capacity;future policy refi nement could draw on international experience to strengthen assessment standardization,broaden eff ective coverage,and improve the sustainability of fi nancing.Conclusions:China can benefi t from international experience in developing standardized assessment protocols,diversifying funding mechanisms,and establishing professional service delivery systems,while acknowledging contextual constraints unique to developing country settings.展开更多
The scaling-up of electrochemical CO_(2)reduction requires circumventing the CO_(2)loss as carbonates under alkaline conditions.Zero-gap MEA cell configurations with a proton exchange membrane represent an alternative...The scaling-up of electrochemical CO_(2)reduction requires circumventing the CO_(2)loss as carbonates under alkaline conditions.Zero-gap MEA cell configurations with a proton exchange membrane represent an alternative solution in a pure acidic system,but the catalyst layer in direct contact with the hydrated proton environment usually leads to H_(2)evolution dominating.Herein,we show that polydimethyldiallyl-ammonium-chloride-coated Ag(Ag@PDDA)electrode exhibits outstanding performance with a FE of 86%,a single-pass conversion of 72%,and a stability of 28 h for CO production in pure-acid MEA compared with ammonium poly(N-methyl-piperidine-co-pterphenyl)decorated Ag(Ag/QAPPT)and cetyltrimethylammonium bromide decorated Ag(Ag/CTAB).The in situ ATR-SEIRAS reveal that PDDA creates a positive charge-rich protective outer layer and an N-rich hybrid inner layer,which not only suppresses the migration of H+during the electrolysis process and blocks the direct contact between H2O and Ag catalyst,but also promotes the generation from CO_(2)to*COOH in a pure-acid system.This work highlights the importance of polyelectrolyte engineering in regulating the electrocatalytic interface and accelerates the development of proton exchange membrane CO_(2)electrolysis.展开更多
Aqueous zinc-ion batteries(AZIBs) have advantages including low economic cost and high safety.Nevertheless,the serious hydrogen evolution reactions(HER) and rampant growth of Zn dendrite hinder their further developme...Aqueous zinc-ion batteries(AZIBs) have advantages including low economic cost and high safety.Nevertheless,the serious hydrogen evolution reactions(HER) and rampant growth of Zn dendrite hinder their further development.Herein,potassium acetate(KAc) additive with cation/anion synergy effect is added into the ZnSO_(4) electrolyte to effectively promote the oriented uniform Zn deposition and suppress side reactions.According to density functional theory calculation and experimental results,CH_(3)COO^(-)(Ac^(-))anions are capable of forming stronger hydrogen bonds with H_(2)O molecules,leading to an expanded electrochemical stability window,reduced the reactivity of H_(2)O,and hence suppressing HER.Meanwhile,Ac-anions can also preferentially adsorb onto the Zn anode,promoting dense deposition towards the(100) crystal plane.Besides,dissociated K^(+) ions serve as electrostatic shielding cations,which significantly promote uniform Zn deposition and prevent dendrite formation.Thus,the Zn||Zn symmetric cell demonstrates an impressive cycle lifespan of 3000 h at 1.0 m A/cm^(2).Furthermore,the Zn||MnO_(2) full battery exhibits superior stability with a capacity retention of 86.95 % at 2.0 A/g after 4000 cycles.Therefore,the cation/anion synergy effect in KAc additive offers a viable solution to address HER and hinder dendrite growth at the interface of Zn anodes.展开更多
In thefield of agriculture,the development of an early warning diagnostic system is essential for timely detection and accurate diagnosis of diseases in rice plants.This research focuses on identifying the plant diseas...In thefield of agriculture,the development of an early warning diagnostic system is essential for timely detection and accurate diagnosis of diseases in rice plants.This research focuses on identifying the plant diseases and detecting them promptly through the advancements in thefield of computer vision.The images obtained from in-field farms are typically with less visual information.However,there is a significant impact on the classification accuracy in the disease diagnosis due to the lack of high-resolution crop images.We propose a novel Reconstructed Disease Aware–Convolutional Neural Network(RDA-CNN),inspired by recent CNN architectures,that integrates image super resolution and classification into a single model for rice plant disease classification.This network takes low-resolution images of rice crops as input and employs the super resolution layers to transform low-resolution images to super-resolution images to recover appearance such as spots,rot,and lesion on different parts of the rice plants.Extensive experimental results indicated that the proposed RDA-CNN method performs well under diverse aspects generating visually pleasing images and outperforms better than other con-ventional Super Resolution(SR)methods.Furthermore,these super-resolution images are subsequently passed through deep classification layers for disease classi-fication.The results demonstrate that the RDA-CNN significantly boosts the clas-sification performance by nearly 4–6%compared with the baseline architectures.展开更多
The data generated from non-Euclidean domains and its graphical representation(with complex-relationship object interdependence)applications has observed an exponential growth.The sophistication of graph data has pose...The data generated from non-Euclidean domains and its graphical representation(with complex-relationship object interdependence)applications has observed an exponential growth.The sophistication of graph data has posed consequential obstacles to the existing machine learning algorithms.In this study,we have considered a revamped version of a semi-supervised learning algorithm for graph-structured data to address the issue of expanding deep learning approaches to represent the graph data.Additionally,the quantum information theory has been applied through Graph Neural Networks(GNNs)to generate Riemannian metrics in closed-form of several graph layers.In further,to pre-process the adjacency matrix of graphs,a new formulation is established to incorporate high order proximities.The proposed scheme has shown outstanding improvements to overcome the deficiencies in Graph Convolutional Network(GCN),particularly,the information loss and imprecise information representation with acceptable computational overhead.Moreover,the proposed Quantum Graph Convolutional Network(QGCN)has significantly strengthened the GCN on semi-supervised node classification tasks.In parallel,it expands the generalization process with a significant difference by making small random perturbationsG of the graph during the training process.The evaluation results are provided on three benchmark datasets,including Citeseer,Cora,and PubMed,that distinctly delineate the superiority of the proposed model in terms of computational accuracy against state-of-the-art GCN and three other methods based on the same algorithms in the existing literature.展开更多
Researchers and scientists need rapid access to text documents such as research papers,source code and dissertations.Many research documents are available on the Internet and need more time to retrieve exact documents...Researchers and scientists need rapid access to text documents such as research papers,source code and dissertations.Many research documents are available on the Internet and need more time to retrieve exact documents based on keywords.An efficient classification algorithm for retrieving documents based on keyword words is required.The traditional algorithm performs less because it never considers words’polysemy and the relationship between bag-of-words in keywords.To solve the above problem,Semantic Featured Convolution Neural Networks(SF-CNN)is proposed to obtain the key relationships among the searching keywords and build a structure for matching the words for retrieving correct text documents.The proposed SF-CNN is based on deep semantic-based bag-of-word representation for document retrieval.Traditional deep learning methods such as Convolutional Neural Network and Recurrent Neural Network never use semantic representation for bag-of-words.The experiment is performed with different document datasets for evaluating the performance of the proposed SF-CNN method.SF-CNN classifies the documents with an accuracy of 94%than the traditional algorithms.展开更多
BACKGROUND:The dynamic monitoring of immune status is crucial to the precise and individualized treatment of sepsis.In this study,we aim to introduce a model to describe and monitor the immune status of sepsis and to ...BACKGROUND:The dynamic monitoring of immune status is crucial to the precise and individualized treatment of sepsis.In this study,we aim to introduce a model to describe and monitor the immune status of sepsis and to explore its prognostic value.METHODS:A prospective observational study was carried out in Zhongshan Hospital,Fudan University,enrolling septic patients admitted between July 2016 and December 2018.Blood samples were collected at days 1 and 3.Serum cytokine levels(e.g.,tumor necrosis factor-α[TNF-α],interleukin-10[IL-10])and CD14+monocyte human leukocyte antigen-D-related(HLA-DR)expression were measured to serve as immune markers.Classifi cation of each immune status,namely systemic inflammatory response syndrome(SIRS),compensatory anti-inflammatory response syndrome(CARS),and mixed antagonistic response syndrome(MARS),was defined based on levels of immune markers.Changes of immune status were classifi ed into four groups which were stabilization(SB),deterioration(DT),remission(RM),and non-remission(NR).RESULTS:A total of 174 septic patients were enrolled including 50 non-survivors.Multivariate analysis discovered that IL-10 and HLA-DR expression levels at day 3 were independent prognostic factors.Patients with MARS had the highest mortality rate.Immune status of 46.1%patients changed from day 1 to day 3.Among four groups of immune status changes,DT had the highest mortality rate,followed by NR,RM,and SB with mortality rates of 64.7%,42.9%,and 11.2%,respectively.CONCLUSIONS:Severe immune disorder defi ned as MARS or deterioration of immune status defi ned as DT lead to the worst outcomes.The preliminary model of the classifi cation and dynamic monitoring of immune status based on immune markers has prognostic values and is worthy of further investigation.展开更多
Skin cancer is usually classified as melanoma and non-melanoma.Melanoma now represents 75%of humans passing away worldwide and is one of the most brutal types of cancer.Previously,studies were not mainly focused on fea...Skin cancer is usually classified as melanoma and non-melanoma.Melanoma now represents 75%of humans passing away worldwide and is one of the most brutal types of cancer.Previously,studies were not mainly focused on feature extraction of Melanoma,which caused the classification accuracy.However,in this work,Histograms of orientation gradients and local binary pat-terns feature extraction procedures are used to extract the important features such as asymmetry,symmetry,boundary irregularity,color,diameter,etc.,and are removed from both melanoma and non-melanoma images.This proposed Effi-cient Classification Systems for the Diagnosis of Melanoma(ECSDM)framework consists of different schemes such as preprocessing,segmentation,feature extrac-tion,and classification.We used Machine Learning(ML)and Deep Learning(DL)classifiers in the classification framework.The ML classifier is Naïve Bayes(NB)and Support Vector Machines(SVM).And also,DL classification frame-work of the Convolution Neural Network(CNN)is used to classify the melanoma and benign images.The results show that the Neural Network(NNET)classifier’achieves 97.17%of accuracy when contrasting with ML classifiers.展开更多
The analysis of remote sensing image areas is needed for climate detec-tion and management,especially for monitoringflood disasters in critical environ-ments and applications.Satellites are mostly used to detect disast...The analysis of remote sensing image areas is needed for climate detec-tion and management,especially for monitoringflood disasters in critical environ-ments and applications.Satellites are mostly used to detect disasters on Earth,and they have advantages in capturing Earth images.Using the control technique,Earth images can be used to obtain detailed terrain information.Since the acquisi-tion of satellite and aerial imagery,this system has been able to detectfloods,and with increasing convenience,flood detection has become more desirable in the last few years.In this paper,a Big Data Set-based Progressive Image Classification Algorithm(PICA)system is introduced to implement an image processing tech-nique,detect disasters,and determine results with the help of the PICA,which allows disaster analysis to be extracted more effectively.The PICA is essential to overcoming strong shadows,for proper access to disaster characteristics to false positives by operators,and to false predictions that affect the impact of the disas-ter.The PICA creates tailoring and adjustments obtained from satellite images before training and post-disaster aerial image data patches.Two types of proposed PICA systems detect disasters faster and more accurately(95.6%).展开更多
Cybercrime has increased considerably in recent times by creating new methods of stealing,changing,and destroying data in daily lives.Portable Docu-ment Format(PDF)has been traditionally utilized as a popular way of s...Cybercrime has increased considerably in recent times by creating new methods of stealing,changing,and destroying data in daily lives.Portable Docu-ment Format(PDF)has been traditionally utilized as a popular way of spreading malware.The recent advances of machine learning(ML)and deep learning(DL)models are utilized to detect and classify malware.With this motivation,this study focuses on the design of mayfly optimization with a deep belief network for PDF malware detection and classification(MFODBN-MDC)technique.The major intention of the MFODBN-MDC technique is for identifying and classify-ing the presence of malware exist in the PDFs.The proposed MFODBN-MDC method derives a new MFO algorithm for the optimal selection of feature subsets.In addition,Adamax optimizer with the DBN model is used for PDF malware detection and classification.The design of the MFO algorithm to select features and Adamax based hyperparameter tuning for PDF malware detection and classi-fication demonstrates the novelty of the work.For demonstrating the improved outcomes of the MFODBN-MDC model,a wide range of simulations are exe-cuted,and the results are assessed in various aspects.The comparison study high-lighted the enhanced outcomes of the MFODBN-MDC model over the existing techniques with maximum precision,recall,and F1 score of 97.42%,97.33%,and 97.33%,respectively.展开更多
Eye state classification acts as a vital part of the biomedical sector,for instance,smart home device control,drowsy driving recognition,and so on.The modifications in the cognitive levels can be reflected via transformi...Eye state classification acts as a vital part of the biomedical sector,for instance,smart home device control,drowsy driving recognition,and so on.The modifications in the cognitive levels can be reflected via transforming the electro-encephalogram(EEG)signals.The deep learning(DL)models automated extract the features and often showcased improved outcomes over the conventional clas-sification model in the recognition processes.This paper presents an Ensemble Deep Learning with Chimp Optimization Algorithm for EEG Eye State Classifi-cation(EDLCOA-ESC).The proposed EDLCOA-ESC technique involves min-max normalization approach as a pre-processing step.Besides,wavelet packet decomposition(WPD)technique is employed for the extraction of useful features from the EEG signals.In addition,an ensemble of deep sparse autoencoder(DSAE)and kernel ridge regression(KRR)models are employed for EEG Eye State classification.Finally,hyperparameters tuning of the DSAE model takes place using COA and thereby boost the classification results to a maximum extent.An extensive range of simulation analysis on the benchmark dataset is car-ried out and the results reported the promising performance of the EDLCOA-ESC technique over the recent approaches with maximum accuracy of 98.50%.展开更多
Nowadays short texts can be widely found in various social data in relation to the 5G-enabled Internet of Things (IoT). Short text classification is a challenging task due to its sparsity and the lack of context. Prev...Nowadays short texts can be widely found in various social data in relation to the 5G-enabled Internet of Things (IoT). Short text classification is a challenging task due to its sparsity and the lack of context. Previous studies mainly tackle these problems by enhancing the semantic information or the statistical information individually. However, the improvement achieved by a single type of information is limited, while fusing various information may help to improve the classification accuracy more effectively. To fuse various information for short text classification, this article proposes a feature fusion method that integrates the statistical feature and the comprehensive semantic feature together by using the weighting mechanism and deep learning models. In the proposed method, we apply Bidirectional Encoder Representations from Transformers (BERT) to generate word vectors on the sentence level automatically, and then obtain the statistical feature, the local semantic feature and the overall semantic feature using Term Frequency-Inverse Document Frequency (TF-IDF) weighting approach, Convolutional Neural Network (CNN) and Bidirectional Gate Recurrent Unit (BiGRU). Then, the fusion feature is accordingly obtained for classification. Experiments are conducted on five popular short text classification datasets and a 5G-enabled IoT social dataset and the results show that our proposed method effectively improves the classification performance.展开更多
基金supported in part by the National Science Fund for Distinguished Young Scholars under Grant (42025403)the National Key Research and Development Plan of China (2021YFA0716800)the National Key Research and Development Plan of China (2022YFC2903804)。
文摘The increasing risk of ground pressure disasters resulting from deep well mining highlights the urgent need for advanced monitoring and early warning systems.Ground pressure monitoring,supported by microseismic technology,plays a pivotal role in ensuring mine safety by enabling real-time identifi cation and accurate classification of vibration signals such as microseismic signals,blasting signals,and noise.These classifications are critical for improving the efficacy of ground pressure monitoring systems,conducting stability analyses of deep rock masses,and implementing timely and precise roadway support measures.Such eff orts are essential for mitigating ground pressure disasters and ensuring safe mining operations.This study proposes an artificial intelligence-based automatic classification network model for mine vibration signals.Based on conventional convolutional neural networks,the proposed model further incorporates long short-term memory(LSTM)networks and attention mechanisms.The LSTM component eff ectively captures temporal correlations in time-series mining vibration data,while the attention mechanism enhances the models’ability to focus on critical features within the data.To validate the eff ectiveness of our proposed model,a dataset comprising 480,526 waveform records collected in 2022 by the microseismic monitoring system at Guangxi Shanhu Tungsten Mine was used for training,validation,and testing purposes.Results demonstrate that the proposed artifi cial intelligence-based classifi cation method achieves a higher recognition accuracy of 92.21%,significantly outperforming traditional manual classification methods.The proposed model represents a signifi cant advancement in ground pressure monitoring and disaster mitigation.
基金financially supported by the Excellent Youth Scholars Program of State Key Laboratory of Complex Nonferrous Metal Resource Clean Utilization,Kunming University of Science and Technology,China(No.YXQN-2024003)the Central Government-Guided Local Science and Technology Development Fund Project,China(No.202407AB110022)。
文摘The rapid advancement of modern science and technology,coupled with the recent surge in new-energy electric vehicles,has significantly boosted the demand for lithium.This has promoted the development and efficient utilization of lepidolite as a lithium source.Therefore,the processes for the flotation of lepidolite have been studied in depth,particularly the development and use of lepidolite flotation collectors and the action mechanism of the collectors on the lepidolite surface.Based on the crystal-structure characteristics of lepidolite minerals,this review focuses on the application of anionic collectors,amine cationic collectors(primary amines,quaternary ammonium salts,ether amines,and Gemini amines),and combined collectors to the flotation behavior of lepidolite as well as the adsorption mechanisms.New directions and technologies for the controllable flotation of lepidolite are proposed,including process improvement,reagent synthesis,and mechanistic research.This analysis demonstrates the need for the further study of the complex environment inside lepidolite and pulp.By using modern analytical detection methods and quantum chemical calculations,research on reagents for the flotation of lepidolite has expanded,providing new concepts and references for the efficient flotation recovery and utilization of lepidolite.
基金supported by the Natural Science Foundation of China Grant No.52272289 and 5240223,and JSPS(Japan Society for the Promotion of Science)of Grant No.22K19088,23H00313,24H02202,and 24H02205。
文摘NiFe-layered double hydroxides(NiFe-LDHs)are among the most promising earth-abundant electrocatalysts for the oxygen evolution reaction(OER)in alkaline media.However,their practical application is hindered by intrinsic activity limitations and poor stability,primarily due to the asymmetric adsorption of oxygen intermediates.To overcome this,the binding strength must be synergistically tuned to a moderate level to optimize catalytic performance.Here,we engineered NiFeCoCr LDH through Co doping to enhance electrical conductivity and controlled Cr leaching to introduce cationic vacancies for modulating intermediate binding strength in NiFe LDH.X-ray absorption near-edge structure and extended X-ray absorption fine structure analyses reveal that NiFe-LDH with Co doping and Cr vacancies modulates the Ni oxidation state and local coordination environment,leading to a balanced electronic structure and enhanced structural complexity around the Ni sites.Additionally,these vacancies can trap OH^(-)/H_(2)O species,which can serve as a reservoir for OH^(-) transfer,facilitating the rapid formation of OER intermediates and enhancing catalytic performance at high current densities.As a result,V_(Cr)-NiFeCo LDH achieves 1.6 A cm^(-2)current density at 1.7 V vs.RHE while maintaining stable operation for over 1000 h at 500 mA cm^(-2).Density functional theory(DFT)calculations validate the synergistic effects of Co doping and Cr-induced vacancies on intermediate binding energies and improved OER kinetics.Overall,this work presents a rational design strategy to simultaneously enhance the activity and durability of NiFe-based OER catalysts for their application in high-performance alkaline water electrolysis.
基金supported by the National Key R&D Program of China(Nos.2023YFC3404200,2023YFC34042012023YFC3404202)+1 种基金the National Natural Science Foundation of China(No.22575253)the Beijing Natural Science Foundation(No.Z220025)。
文摘The development of synthetic hybrid biological systems integrating photosynthetic organisms with organic-abiotic functional materials holds significant promise for enhancing photosynthetic processes.The artificial regulation of the state transition between photosystem I(PSI)and photosystem II(PSII)represents a strategic and promising approach for improving the efficiency of natural photosynthesis.In this study,we demonstrate that poly(benzimidazolium-phenylthiophene)(CP4)featuring a flexible cationic backbone exhibits superior ultraviolet light-harvesting capability.The polymer CP4 enhanced PSI activity in Chlorella pyrenoidosa(C.pyrenoidosa),subsequently promoting PSII activity and augmenting overall photosynthetic performance.During light-dependent reactions,CP4 significantly accelerated photosynthetic electron transfer,resulting in a 330%increase in the production of oxygen and 93%and 96%increases in the ATP and NADPH contents,respectively.In the context of dark reactions,CP4 facilitated the conversion and utilization of light energy,leading to a 6%increase in both carbohydrate and protein contents.These findings indicate that synthetic light-harvesting polymer materials exhibit considerable application potential in the field of biomass production through enhancement of natural photosynthetic efficiency.
文摘Cataract is the leading cause of reversible blindness worldwide,affecting millions,particularly the elderly.Over 65 million people suffer from significant visual impairment due to cataracts,with the burden being highest in low-and middle-income countries where access to surgery is limited.Cataract surgery,one of the most commonly performed and cost-effective procedures,has evolved significantly.Traditional extracapsular cataract extraction(ECCE)has been largely replaced by phacoemulsifi cation,which uses ultrasonic energy through a small incision,reducing recovery time and complications.More recently,femtosecond laser-assisted cataract surgery(FLACS)has emerged,off ering enhanced precision but with ongoing evaluation of its cost-eff ectiveness.Intraocular lenses(IOLs)now allow for customized visual outcomes,addressing distance,near,and intermediate vision.Despite its safety,cataract surgery can still result in complications such as corneal edema and posterior capsular opacifi cation,requiring careful surgical management and patient education.
基金supported by the National Key R&D Program of China(2023YFB2406000)the National Natural Science Foundation of China(22479057,52172201,51732005)。
文摘Sodium layered oxides stand out as one of the most promising cathodes for sodium-ion batteries due to their high energy density,elemental abundance,and scalability.However,their practical applications are restricted by interplanar gliding,cation migration,and the formation of intragranular microcracks,which collectively lead to rapid structural degradation and capacity loss.Herein,we rationally design an ultrastable O3-type Na_(0.94)Ca_(0.03)Ni_(1/3)Fe_(1/3)Mn_(1/3)O_(2) cathode,in which Ca^(2+)cations act as pillars within the NaO_(2)slabs,suppressing the irreversible phase transitions and Na/TM cation migration commonly observed in layered oxides.Multiscale in situ and ex situ techniques,combined with post-mortem analysis,reveal that the Ca-pillared pinning effect not only effectively suppresses the interplanar gliding and stress accumulation within the crystal phase but also restrains Na/TM cation migration and surface reconstruction in near-surface regions.Benefiting from the combined effects of structural stabilization,the Ca-pillared cathode exhibits a superior cycling stability,retaining 81.6%of its capacity after 1000 cycles at 2 C within the voltage range of 2.0-4.0 V,along with significantly enhanced wide-temperature(from-40 to 80℃)performance.This work highlights another critical role of Ca pillars in suppressing cation migration and surface structural degradation beyond preventing adverse interplanar gliding,offering valuable insights for designing long-life and wide-temperature layered oxide cathodes.
基金financially supported the National Key R&D Program of China (No.2022YFA1502902)the National Natural Science Foundation of China (NSFC,Nos.22475152 and U21A20286)the 111 Project of China (No.D17003)。
文摘Converting CO_(2) into methanol(CH_(3)OH),a high-value-added liquid-phase product,through efficient and highly selective photocatalysis remains a significant challenge.Herein,we present a straightforward cation exchange strategy for the in-situ growth of BiVO_(4) on an InVO_(4) substrate to generate a Z-scheme heterojunction of InVO_(4)/BiVO_(4) .This in-situ partial transformation approach endows the InVO_(4)/BiVO_(4) heterojunction with a tightly connected interface,resulting in a significant improvement in charge separation efficiency between InVO_(4) and BiVO_(4).Moreover,the construction of the heterojunction reduces the formation energy barrier of the ^(*)COOH intermediate during the photoreduction of CO_(2) and increases the desorption energy barrier of the ^(*)CO intermediate,facilitating the deep reduction of ^(*)CO.Consequently,the InVO_(4)/BiVO_(4) heterojunction is capable of photocatalytic CO_(2) reduction to CH_(3)OH with high efficiency and selectivity.Under conditions where water serves as the electron source and a light intensity of 100 m W/cm^(2),the yield of CH_(3)OH reaches 130.5 μmol g^(-1)h^(-1) with a selectivity of 92 %,outperforming photocatalysts reported under similar conditions.
基金funded by the China Disabled Persons’Federation under its 2024 research project(Grant No.2024CDPFAT-47)the Yancheng Social Science Foundation(Grant No.25skB252).
文摘Background:Home accessibility modifi cations are crucial for promoting independent living and quality of life among persons with disabilities.While developed countries have established comprehensive policy frameworks,developing nations like China face unique challenges in program design and implementation.Objective:This study conducts a systematic comparative analysis of home accessibility modification policies across China,Japan,Germany,and Sweden,identifying key policy dimensions and proposing evidence-based recommendations for strengthening China’s policy framework.Methods:We employed a multi-dimensional analytical framework examining legislative foundations,eligibility criteria,funding mechanisms,and service delivery models.Data were collected from primary legislation,governmental regulations,official statistics,and peer-reviewed literature.Results:Significant cross-national variations exist in policy approaches.Japan and Germany utilize social insurance models with standardized assessments,Sweden adopts a universal rights-based approach,while China employs a targeted assistance model focused on economically disadvantaged households.China completed 1.28 million household renovations during its 14th Five-Year Plan,demonstrating strong implementation capacity;future policy refi nement could draw on international experience to strengthen assessment standardization,broaden eff ective coverage,and improve the sustainability of fi nancing.Conclusions:China can benefi t from international experience in developing standardized assessment protocols,diversifying funding mechanisms,and establishing professional service delivery systems,while acknowledging contextual constraints unique to developing country settings.
基金financial support of the National Natural Science Foundation of China(NSFC)(52394202,52021004,52301232,and 52476056)the Natural Science Foundation of Chongqing Province(2024NSCQ-MSX1109).
文摘The scaling-up of electrochemical CO_(2)reduction requires circumventing the CO_(2)loss as carbonates under alkaline conditions.Zero-gap MEA cell configurations with a proton exchange membrane represent an alternative solution in a pure acidic system,but the catalyst layer in direct contact with the hydrated proton environment usually leads to H_(2)evolution dominating.Herein,we show that polydimethyldiallyl-ammonium-chloride-coated Ag(Ag@PDDA)electrode exhibits outstanding performance with a FE of 86%,a single-pass conversion of 72%,and a stability of 28 h for CO production in pure-acid MEA compared with ammonium poly(N-methyl-piperidine-co-pterphenyl)decorated Ag(Ag/QAPPT)and cetyltrimethylammonium bromide decorated Ag(Ag/CTAB).The in situ ATR-SEIRAS reveal that PDDA creates a positive charge-rich protective outer layer and an N-rich hybrid inner layer,which not only suppresses the migration of H+during the electrolysis process and blocks the direct contact between H2O and Ag catalyst,but also promotes the generation from CO_(2)to*COOH in a pure-acid system.This work highlights the importance of polyelectrolyte engineering in regulating the electrocatalytic interface and accelerates the development of proton exchange membrane CO_(2)electrolysis.
基金financially supported by the National Natural Science Foundation of China (No.52372188)the 111 Project (No.D17007)2023 Introduction of studying abroad talent program。
文摘Aqueous zinc-ion batteries(AZIBs) have advantages including low economic cost and high safety.Nevertheless,the serious hydrogen evolution reactions(HER) and rampant growth of Zn dendrite hinder their further development.Herein,potassium acetate(KAc) additive with cation/anion synergy effect is added into the ZnSO_(4) electrolyte to effectively promote the oriented uniform Zn deposition and suppress side reactions.According to density functional theory calculation and experimental results,CH_(3)COO^(-)(Ac^(-))anions are capable of forming stronger hydrogen bonds with H_(2)O molecules,leading to an expanded electrochemical stability window,reduced the reactivity of H_(2)O,and hence suppressing HER.Meanwhile,Ac-anions can also preferentially adsorb onto the Zn anode,promoting dense deposition towards the(100) crystal plane.Besides,dissociated K^(+) ions serve as electrostatic shielding cations,which significantly promote uniform Zn deposition and prevent dendrite formation.Thus,the Zn||Zn symmetric cell demonstrates an impressive cycle lifespan of 3000 h at 1.0 m A/cm^(2).Furthermore,the Zn||MnO_(2) full battery exhibits superior stability with a capacity retention of 86.95 % at 2.0 A/g after 4000 cycles.Therefore,the cation/anion synergy effect in KAc additive offers a viable solution to address HER and hinder dendrite growth at the interface of Zn anodes.
文摘In thefield of agriculture,the development of an early warning diagnostic system is essential for timely detection and accurate diagnosis of diseases in rice plants.This research focuses on identifying the plant diseases and detecting them promptly through the advancements in thefield of computer vision.The images obtained from in-field farms are typically with less visual information.However,there is a significant impact on the classification accuracy in the disease diagnosis due to the lack of high-resolution crop images.We propose a novel Reconstructed Disease Aware–Convolutional Neural Network(RDA-CNN),inspired by recent CNN architectures,that integrates image super resolution and classification into a single model for rice plant disease classification.This network takes low-resolution images of rice crops as input and employs the super resolution layers to transform low-resolution images to super-resolution images to recover appearance such as spots,rot,and lesion on different parts of the rice plants.Extensive experimental results indicated that the proposed RDA-CNN method performs well under diverse aspects generating visually pleasing images and outperforms better than other con-ventional Super Resolution(SR)methods.Furthermore,these super-resolution images are subsequently passed through deep classification layers for disease classi-fication.The results demonstrate that the RDA-CNN significantly boosts the clas-sification performance by nearly 4–6%compared with the baseline architectures.
基金supported by the National Key Research and Development Program of China(2018YFB1600600)the National Natural Science Foundation of China under(61976034,U1808206)the Dalian Science and Technology Innovation Fund(2019J12GX035).
文摘The data generated from non-Euclidean domains and its graphical representation(with complex-relationship object interdependence)applications has observed an exponential growth.The sophistication of graph data has posed consequential obstacles to the existing machine learning algorithms.In this study,we have considered a revamped version of a semi-supervised learning algorithm for graph-structured data to address the issue of expanding deep learning approaches to represent the graph data.Additionally,the quantum information theory has been applied through Graph Neural Networks(GNNs)to generate Riemannian metrics in closed-form of several graph layers.In further,to pre-process the adjacency matrix of graphs,a new formulation is established to incorporate high order proximities.The proposed scheme has shown outstanding improvements to overcome the deficiencies in Graph Convolutional Network(GCN),particularly,the information loss and imprecise information representation with acceptable computational overhead.Moreover,the proposed Quantum Graph Convolutional Network(QGCN)has significantly strengthened the GCN on semi-supervised node classification tasks.In parallel,it expands the generalization process with a significant difference by making small random perturbationsG of the graph during the training process.The evaluation results are provided on three benchmark datasets,including Citeseer,Cora,and PubMed,that distinctly delineate the superiority of the proposed model in terms of computational accuracy against state-of-the-art GCN and three other methods based on the same algorithms in the existing literature.
文摘Researchers and scientists need rapid access to text documents such as research papers,source code and dissertations.Many research documents are available on the Internet and need more time to retrieve exact documents based on keywords.An efficient classification algorithm for retrieving documents based on keyword words is required.The traditional algorithm performs less because it never considers words’polysemy and the relationship between bag-of-words in keywords.To solve the above problem,Semantic Featured Convolution Neural Networks(SF-CNN)is proposed to obtain the key relationships among the searching keywords and build a structure for matching the words for retrieving correct text documents.The proposed SF-CNN is based on deep semantic-based bag-of-word representation for document retrieval.Traditional deep learning methods such as Convolutional Neural Network and Recurrent Neural Network never use semantic representation for bag-of-words.The experiment is performed with different document datasets for evaluating the performance of the proposed SF-CNN method.SF-CNN classifies the documents with an accuracy of 94%than the traditional algorithms.
基金the National Natural Science Foundation of China(81471840,81171837)the Shanghai Traditional Medicine Development Project(ZY3-CCCX3-3018,ZHYY-ZXYJH-201615)the Key Project of Shanghai Municipal Health Bureau(2016ZB0202).
文摘BACKGROUND:The dynamic monitoring of immune status is crucial to the precise and individualized treatment of sepsis.In this study,we aim to introduce a model to describe and monitor the immune status of sepsis and to explore its prognostic value.METHODS:A prospective observational study was carried out in Zhongshan Hospital,Fudan University,enrolling septic patients admitted between July 2016 and December 2018.Blood samples were collected at days 1 and 3.Serum cytokine levels(e.g.,tumor necrosis factor-α[TNF-α],interleukin-10[IL-10])and CD14+monocyte human leukocyte antigen-D-related(HLA-DR)expression were measured to serve as immune markers.Classifi cation of each immune status,namely systemic inflammatory response syndrome(SIRS),compensatory anti-inflammatory response syndrome(CARS),and mixed antagonistic response syndrome(MARS),was defined based on levels of immune markers.Changes of immune status were classifi ed into four groups which were stabilization(SB),deterioration(DT),remission(RM),and non-remission(NR).RESULTS:A total of 174 septic patients were enrolled including 50 non-survivors.Multivariate analysis discovered that IL-10 and HLA-DR expression levels at day 3 were independent prognostic factors.Patients with MARS had the highest mortality rate.Immune status of 46.1%patients changed from day 1 to day 3.Among four groups of immune status changes,DT had the highest mortality rate,followed by NR,RM,and SB with mortality rates of 64.7%,42.9%,and 11.2%,respectively.CONCLUSIONS:Severe immune disorder defi ned as MARS or deterioration of immune status defi ned as DT lead to the worst outcomes.The preliminary model of the classifi cation and dynamic monitoring of immune status based on immune markers has prognostic values and is worthy of further investigation.
文摘Skin cancer is usually classified as melanoma and non-melanoma.Melanoma now represents 75%of humans passing away worldwide and is one of the most brutal types of cancer.Previously,studies were not mainly focused on feature extraction of Melanoma,which caused the classification accuracy.However,in this work,Histograms of orientation gradients and local binary pat-terns feature extraction procedures are used to extract the important features such as asymmetry,symmetry,boundary irregularity,color,diameter,etc.,and are removed from both melanoma and non-melanoma images.This proposed Effi-cient Classification Systems for the Diagnosis of Melanoma(ECSDM)framework consists of different schemes such as preprocessing,segmentation,feature extrac-tion,and classification.We used Machine Learning(ML)and Deep Learning(DL)classifiers in the classification framework.The ML classifier is Naïve Bayes(NB)and Support Vector Machines(SVM).And also,DL classification frame-work of the Convolution Neural Network(CNN)is used to classify the melanoma and benign images.The results show that the Neural Network(NNET)classifier’achieves 97.17%of accuracy when contrasting with ML classifiers.
基金funded by Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia,under grant No.(PNURSP2022R161).
文摘The analysis of remote sensing image areas is needed for climate detec-tion and management,especially for monitoringflood disasters in critical environ-ments and applications.Satellites are mostly used to detect disasters on Earth,and they have advantages in capturing Earth images.Using the control technique,Earth images can be used to obtain detailed terrain information.Since the acquisi-tion of satellite and aerial imagery,this system has been able to detectfloods,and with increasing convenience,flood detection has become more desirable in the last few years.In this paper,a Big Data Set-based Progressive Image Classification Algorithm(PICA)system is introduced to implement an image processing tech-nique,detect disasters,and determine results with the help of the PICA,which allows disaster analysis to be extracted more effectively.The PICA is essential to overcoming strong shadows,for proper access to disaster characteristics to false positives by operators,and to false predictions that affect the impact of the disas-ter.The PICA creates tailoring and adjustments obtained from satellite images before training and post-disaster aerial image data patches.Two types of proposed PICA systems detect disasters faster and more accurately(95.6%).
文摘Cybercrime has increased considerably in recent times by creating new methods of stealing,changing,and destroying data in daily lives.Portable Docu-ment Format(PDF)has been traditionally utilized as a popular way of spreading malware.The recent advances of machine learning(ML)and deep learning(DL)models are utilized to detect and classify malware.With this motivation,this study focuses on the design of mayfly optimization with a deep belief network for PDF malware detection and classification(MFODBN-MDC)technique.The major intention of the MFODBN-MDC technique is for identifying and classify-ing the presence of malware exist in the PDFs.The proposed MFODBN-MDC method derives a new MFO algorithm for the optimal selection of feature subsets.In addition,Adamax optimizer with the DBN model is used for PDF malware detection and classification.The design of the MFO algorithm to select features and Adamax based hyperparameter tuning for PDF malware detection and classi-fication demonstrates the novelty of the work.For demonstrating the improved outcomes of the MFODBN-MDC model,a wide range of simulations are exe-cuted,and the results are assessed in various aspects.The comparison study high-lighted the enhanced outcomes of the MFODBN-MDC model over the existing techniques with maximum precision,recall,and F1 score of 97.42%,97.33%,and 97.33%,respectively.
基金Supported by A VIDI grant from the Netherlands Organiza-tion for Scientific Research(NWO,to Weersma RK),No.016.136.308an AGIKO grant from the Netherlands Organiza-tion for Scientific Research(NWO to Visschedijk MC),No.92.003.577MLDS grant of the Dutch Digestive Foundation,No.WO 11-72(to Alberts R)
文摘AIM: To validate the Montreal classification system for Crohn’s disease (CD) and ulcerative colitis (UC) within the Netherlands.
基金supported by the Researchers Supporting Program(TUMA-Project-2021–27)Almaarefa University,Riyadh,Saudi ArabiaTaif University Researchers Supporting Project Number(TURSP-2020/161),Taif University,Taif,Saudi Arabia.
文摘Eye state classification acts as a vital part of the biomedical sector,for instance,smart home device control,drowsy driving recognition,and so on.The modifications in the cognitive levels can be reflected via transforming the electro-encephalogram(EEG)signals.The deep learning(DL)models automated extract the features and often showcased improved outcomes over the conventional clas-sification model in the recognition processes.This paper presents an Ensemble Deep Learning with Chimp Optimization Algorithm for EEG Eye State Classifi-cation(EDLCOA-ESC).The proposed EDLCOA-ESC technique involves min-max normalization approach as a pre-processing step.Besides,wavelet packet decomposition(WPD)technique is employed for the extraction of useful features from the EEG signals.In addition,an ensemble of deep sparse autoencoder(DSAE)and kernel ridge regression(KRR)models are employed for EEG Eye State classification.Finally,hyperparameters tuning of the DSAE model takes place using COA and thereby boost the classification results to a maximum extent.An extensive range of simulation analysis on the benchmark dataset is car-ried out and the results reported the promising performance of the EDLCOA-ESC technique over the recent approaches with maximum accuracy of 98.50%.
基金supported in part by the Beijing Natural Science Foundation under grants M21032 and 19L2029in part by the National Natural Science Foundation of China under grants U1836106 and 81961138010in part by the Scientific and Technological Innovation Foundation of Foshan under grants BK21BF001 and BK20BF010.
文摘Nowadays short texts can be widely found in various social data in relation to the 5G-enabled Internet of Things (IoT). Short text classification is a challenging task due to its sparsity and the lack of context. Previous studies mainly tackle these problems by enhancing the semantic information or the statistical information individually. However, the improvement achieved by a single type of information is limited, while fusing various information may help to improve the classification accuracy more effectively. To fuse various information for short text classification, this article proposes a feature fusion method that integrates the statistical feature and the comprehensive semantic feature together by using the weighting mechanism and deep learning models. In the proposed method, we apply Bidirectional Encoder Representations from Transformers (BERT) to generate word vectors on the sentence level automatically, and then obtain the statistical feature, the local semantic feature and the overall semantic feature using Term Frequency-Inverse Document Frequency (TF-IDF) weighting approach, Convolutional Neural Network (CNN) and Bidirectional Gate Recurrent Unit (BiGRU). Then, the fusion feature is accordingly obtained for classification. Experiments are conducted on five popular short text classification datasets and a 5G-enabled IoT social dataset and the results show that our proposed method effectively improves the classification performance.
