Feature Embedding Research Articles

Contrastive-learning of language embedding and biological features for cross modality encoding and effector prediction

Identifying and characterizing virulence proteins secreted by Gram-negative bacteria are fundamental for deciphering microbial pathogenicity as well as aiding the development of therapeutic strategies. Effector predictors utilizing pre-trained protein language models (PLMs) have shown sound performance by leveraging extensive evolutionary and sequential protein features. However, the accuracy and sensitivity of effector prediction remain challenging. Here, we introduce a model named Contrastive-learning of Language Embedding and Biological Features (CLEF) leveraging contrastive learning to integrate PLM representations with supplementary biological features. Biologically information is captured in learned contextualized embeddings to yield meaningful representations. With cross-modality biological features, CLEF outperforms state-of-the-art (SOTA) models in predicting type III, type IV, and type VI secreted effectors (T3SEs/T4SEs/T6SEs) in enteric pathogens. All experimentally verified effectors in Enterohemorrhagic Escherichia coli and 41 of 43 experimentally verified T3SEs of Salmonella Typhimurium are recognized. Moreover, 12 predicted T3SEs and 11 predicted T6SEs are validated by extensive experiments in Edwardsiella piscicida. Furthermore, integrating omics data via CLEF framework enhances protein representations to illustrate effector-effector interactions and determine in vivo colonization-essential genes. Collectively, CLEF provides a blueprint to bridge the gap between in silico PLM’s capacity and experimental biological information to fulfill complicated tasks.

Multi-channel spatio-temporal graph attention contrastive network for brain disease diagnosis.

Dynamic brain networks (DBNs) can capture the intricate connections and temporal evolution among brain regions, becoming increasingly crucial in the diagnosis of neurological disorders. However, most existing researches tend to focus on isolated brain network sequence segmented by sliding windows, and they are difficult to effectively uncover the higher-order spatio-temporal topological pattern in DBNs. Meantime, it remains a challenge to utilize the structure connectivity prior in the DBNs analysis. To address these problems, we propose a multi-channel spatio-temporal graph attention contrastive network for DBNs analysis. Specifically, we first construct dynamic brain functional networks from fMRI data with sliding windows, and embed the structural connectivity derived from diffusion tensor imaging (DTI) to the dynamic functional connectivity graph representation to construct multi-modal brain network. Second, we develop a multi-channel spatial attention contrastive network to extract topological features from the brain network within each time window. This network incorporates an intra-window graph contrastive constraint to enhance the discriminative ability of the extracted features. Moreover, temporal dependencies across windows are captured by integrating feature embeddings through a self-attention mechanism, and the inter-window recurrent contrastive constraint is devised to extract higher-order spatio-temporal topological features. Finally, a multi-layer perceptron (MLP) is used to classify the brain networks. Experiments on epilepsy and ADNI datasets show that our method outperforms several state-of-the-art approaches in diagnosing performance, and it provides discriminative graph features for related brain diseases.

A novel graph neural network based approach for influenza-like illness nowcasting: exploring the interplay of temporal, geographical, and functional spatial features

BackgroundAccurate and timely monitoring of influenza prevalence is essential for effective healthcare interventions. This study proposes a graph neural network (GNN)-based method to address the issue of cross-regional connectivity in predicting influenza outbreaks, aiming to achieve real-time and accurate influenza prediction.MethodsWe proposed a GNN-based approach with dual topology processing, capturing both geographical and socio-economic associations among counties/cities. The model inputs consist of weekly matrices of influenza-like illness (ILI) rates at city level, along with geographical topology and functional topology. The model construction involves temporal feature extraction through 1-dimensional gated causal convolution, spatial feature embedding through graph convolution, and additional adjustments to enhance spatiotemporal interaction exploration. Evaluation metrics include four commonly used measures: root mean square error (RMSE), mean absolute percentage error (MAPE), mean absolute error (MAE), and Pearson correlation (Corr).ResultsOur approach for predicting influenza outbreaks achieves competitive performance on real-world datasets (Corr = 0.8202; RMSE = 0.0017; MAE = 0.0013; MAPE = 0.0966), surpassing established baselines. Notably, our approach exhibits excellent capability in accurately and timely capturing short-term influenza outbreaks during the flu season, outperforming competitors across all evaluation metrics.ConclusionThe incorporation of dual topology processing and the subsequent fusion mechanism allows the model to explore in-depth spatiotemporal feature interactions. Demonstrating superior performance, our approach shows great potential in early detection of flu trends for facilitating public health decisions and resource optimization.

Efficient deep neural network training via decreasing precision with layer capacity

Low-precision training has emerged as a practical approach, saving the cost of time, memory, and energy during deep neural networks (DNNs) training. Typically, the use of lower precision introduces quantization errors that need to be minimized to maintain model performance, often neglecting to consider the potential benefits of reducing training precision. This paper rethinks low-precision training, highlighting the potential benefits of lowering precision: (1) low precision can serve as a form of regularization in DNN training by constraining excessive variance in the model; (2) layer-wise low precision can be seen as an alternative dimension of sparsity, orthogonal to pruning, contributing to improved generalization in DNNs. Based on these analyses, we propose a simple yet powerful technique–DPC (Decreasing Precision with layer Capacity), which directly assigns different bit-widths to model layers, without the need for an exhaustive analysis of the training process or any delicate low-precision criteria. Thorough extensive experiments on five datasets and fourteen models across various applications consistently demonstrate the effectiveness of the proposed DPC technique in saving computational cost (−16.21%–−44.37%) while achieving comparable or even superior accuracy (up to +0.68%, +0.21% on average). Furthermore, we offer feature embedding visualizations and conduct further analysis with experiments to investigate the underlying mechanisms behind DPC’s effectiveness, enhancing our understanding of low-precision training. Our source code will be released upon paper acceptance.

Joint embedding–classifier learning for interpretable collaborative filtering

BackgroundInterpretability is a topical question in recommender systems, especially in healthcare applications. An interpretable classifier quantifies the importance of each input feature for the predicted item-user association in a non-ambiguous fashion.ResultsWe introduce the novel Joint Embedding Learning-classifier for improved Interpretability (JELI). By combining the training of a structured collaborative-filtering classifier and an embedding learning task, JELI predicts new user-item associations based on jointly learned item and user embeddings while providing feature-wise importance scores. Therefore, JELI flexibly allows the introduction of priors on the connections between users, items, and features. In particular, JELI simultaneously (a) learns feature, item, and user embeddings; (b) predicts new item-user associations; (c) provides importance scores for each feature. Moreover, JELI instantiates a generic approach to training recommender systems by encoding generic graph-regularization constraints.ConclusionsFirst, we show that the joint training approach yields a gain in the predictive power of the downstream classifier. Second, JELI can recover feature-association dependencies. Finally, JELI induces a restriction in the number of parameters compared to baselines in synthetic and drug-repurposing data sets.

Contrastive Dual-Pool Feature Adaption for Domain Incremental Remote Sensing Scene Classification

Remote sensing image classification has achieved remarkable success in environmental monitoring and urban planning using deep neural networks (DNNs). However, the performance of these models is significantly impacted by domain shifts due to seasonal changes, varying atmospheric conditions, and different geographical locations. Existing solutions, including rehearsal-based and prompt-based methods, face limitations such as data privacy concerns, high computational overhead, and unreliable feature embeddings due to domain gaps. To address these challenges, we propose DACL (dual-pool architecture with contrastive learning), a novel framework for domain incremental learning in remote sensing image classification. DACL introduces three key components: (1) a dual-pool architecture comprising a prompt pool for domain-specific tokens and an adapter pool for feature adaptation, enabling efficient domain-specific feature extraction; (2) a dual loss mechanism that combines image-attracting loss and text-separating loss to enhance intra-domain feature discrimination while maintaining clear class boundaries; and (3) a K-means-based domain selector that efficiently matches unknown domain features with existing domain representations using cosine similarity. Our approach eliminates the need for storing historical data while maintaining minimal computational overhead. Extensive experiments on six widely used datasets demonstrate that DACL consistently outperforms state-of-the-art methods in domain incremental learning for remote sensing image classification scenarios, achieving an average accuracy improvement of 4.07% over the best baseline method.

User Behavior Analysis and Security Strategy Optimization of College English Learning Platform Using Dynamic Graph Convolutional Network

This paper aimed to address the shortcomings of current college English learning platforms in user behavior analysis and security protection, especially the lack of accurate abnormal behavior detection and dynamic identity authentication measures. By adopting DGCN (Dynamic Graph Convolutional Network) and XGBoost (eXtreme Gradient Boosting) classification models for abnormal behavior detection, a dynamic identity authentication optimization strategy based on behavior risk scoring is proposed to improve the security of the platform. The user embedding features output from DGCN are input into XGBoost for user behavior classification. Based on the predicted probability of the XGBoost classification model, a risk score is generated for each user, and different identity authentication strategies are defined according to the risk score range. Experimental results show that by modeling multi-layer dynamic graph structures, the accuracy, precision, recall and F1 score of the XGBoost model reached 97.6%, 97.5%, 95.3%, and 96.4% respectively. In addition, when the step size of the DGCN model increases from 1 to 10, its accuracy only slightly decreases from 97.6% to 97.5%, and the user behavior classification performance is stable. The incidence of account theft with dynamic identity authentication is maintained between 3% and 8%, and the average authentication time is increased by 2.1 seconds compared with the traditional password authentication method. The proposed DGCN can optimize the security strategy of the college English learning platform based on the user behavior risk score, and it can select targeted identity authentication strategies according to the user's risk score.

Inductive transfer-learning of high-fidelity aerodynamics from inviscid panel methods

Building accurate and generalizable machine-learning models requires large training datasets. In aerodynamics, quantities of interest are typically governed by complex, non-linear mechanisms in which neural networks are well-suited to address. However, the acquisition of large, high-fidelity datasets from either simulations or experiments can be expensive. In this work, a transfer-learning framework is explored to reduce the reliance on these expensive datasets by exploiting the cost-effectiveness of low-fidelity analyses in constructing extensive datasets, such as the inviscid panel method. By first developing robust base networks from inviscid distributions, target networks can “learn” by simply transferring relevant embedded features to facilitate the modelling of high-fidelity distributions, instead of solely relying on its access to high-fidelity samples. Assessment of the framework reveals performance gains over conventional training schemes in (1) fidelity enhancement from inviscid to high-fidelity pressure distributions; (2) generalizing prior knowledge to learn adjacent skin friction properties even without a low-fidelity equivalent; (3) extrapolation to yet-to-be seen operating conditions. Under conditions of limited high-fidelity samples, test MSE evaluations can be improved by magnitudes of up to 102, 101, and 102 for the three respective tasks. As such, these findings motivate further investigations to support data-scarce surrogate modelling in more empirical settings.

NRGCNMDA: Microbe-Drug Association Prediction Based on Residual Graph Convolutional Networks and Conditional Random Fields.

The process of discovering new drugs related to microbes through traditional biological methods is lengthy and costly. In response to these issues, a new computational model (NRGCNMDA) is proposed to predict microbe-drug associations. First, Node2vec is used to extract potential associations between microorganisms and drugs, and a heterogeneous network of microbes and drugs is constructed. Then, a Graph Convolutional Network incorporating a fusion residual network mechanism (REGCN) is utilized to learn meaningful high-order similarity features. In addition, conditional random fields (CRF) are applied to ensure that microbes and drugs have similar feature embeddings. Finally, unobserved microbe-drug associations are scored based on combined embeddings. The experimental findings demonstrate that the NRGCNMDA approach outperforms several existing deep learning methods, and its AUC and AUPR values are 95.16% and 93.02%, respectively. The case study demonstrates that NRGCNMDA accurately predicts drugs associated with Enterococcus faecalis and Listeria monocytogenes, as well as microbes associated with ibuprofen and tetracycline.

STGLR: A Spacecraft Anomaly Detection Method Based on Spatio-Temporal Graph Learning.

Anomalies frequently occur during the operation of spacecraft in orbit, and studying anomaly detection methods is crucial to ensure the normal operation of spacecraft. Due to the complexity of spacecraft structures, telemetry data possess characteristics such as high dimensionality, complexity, and large scale. Existing methods frequently ignore or fail to explicitly extract the correlation between variables, and due to the lack of prior knowledge, it is difficult to obtain the initial relationship of variables. To address these issues, this paper proposes a new method, namely spatio-temporal graph learning reconstruction (STGLR), for spacecraft anomaly detection. STGLR employs a dynamic graph learning module to infer the initial relationships among telemetry variables. It then constructs a spatio-temporal feature extraction module to capture complex spatio-temporal dependencies among variables, leveraging a graph sample and aggregation network to learn embedded features and incorporating an attention mechanism to adaptively select salient features. Finally, a reconstruction module is used to learn the latent representations of features, capturing the normal patterns in telemetry data and achieving anomaly detection. To validate the effectiveness of the proposed method, experiments were conducted on two public spacecraft datasets, and the results demonstrate that the performance of the STGLR method surpasses existing anomaly detection methods, with an average F1 score exceeding 0.97.

Enhanced Traffic Congestion Prediction Using Attention-Based Multi-Layer GRU Model with Feature Embedding

Enhanced Traffic Congestion Prediction Using Attention-Based Multi-Layer GRU Model with Feature Embedding

Leveraging neighborhood distance awareness for entity alignment in temporal knowledge graphs.

Entity alignment (EA) is a typical strategy for knowledge graph integration, aiming to identify and align different entity pairs representing the same real object from different knowledge graphs. Temporal Knowledge Graph (TKG) extends the static knowledge graph by introducing timestamps. However, since temporal knowledge graphs are constructed based on their own data sources, this usually leads to problems such as missing or redundant entity information in the temporal knowledge graph. Therefore, temporal knowledge graph entity alignment is a meaningful task to avoid the above problems. However, most existing methods tend to ignore the impact of direct and indirect neighborhood information on EA. Therefore, we propose a temporal knowledge graph entity alignment method with neighborhood distance awareness, namely Tem-DA. Tem-DA models direct and indirect neighbors separately, capturing directly connected neighbors with related entities and indirectly connected neighbors with unrelated entities through a distance detection module. Furthermore, we propose a gating mechanism weight the elements in the embedding matrix and through cross-entropy loss function with regularization terms to address adaptive fusion. This mechanism dynamically adjusts the weights of different feature embeddings, thus providing a more flexible and adaptive feature fusion strategy compared with traditional linear weight adjustment methods. Tem-DA effectively captures the temporal information of entities and take advantage of overlapping properties that may have multiple identical time intervals between different entities to encode the temporal information. For some entities that may lack sufficient temporal information, Tem-DA estimates the temporal characteristics of adjacent information and generates temporal embedding vectors. Experimental results on two monolingual TKG datasets show that Tem-DA outperforms popular baseline methods.

Development of face image recognition algorithm using CNN in airport security checkpoints for terrorist early detection

Ensuring airport security is of paramount importance to safeguard the lives of passengers and prevent acts of terrorism. In this context, developing advanced technology for early terrorist detection is crucial. This paper presents a novel approach to enhancing security measures at airport checkpoints by applying Convolutional Neural Network (CNN) and Artificial Neural Network (ANN) algorithms in face image recognition. Our system utilizes state-of-the-art artificial intelligence techniques to analyze facial features. Our research uses VGG architecture and pre-trained with face data as a CNN model. This model is used to extract face embedding features from the dataset. These embedding features are then compressed with Principal Component Analysis (PCA) to obtain the meaningful feature as training data for the ANN algorithm. We trained our system using data from 500 identities data with 60 data for each identity. This training enables our system to recognize known terrorists and individuals on watchlists by comparing the facial features of individuals passing through security checkpoints with those in the database. The proposed CNN-ANN-based face recognition system not only enhances airport security but also significantly reduces the processing time for security checks. It can quickly identify potential threats, allowing security personnel to take appropriate actions in real time ensuring a rapid response to security concerns. We present the architecture, training methodology, and evaluation of the CNN-ANN model, achieving a high accuracy of 91.16% and precision of 91.36%. Through this research, we aim to increase airport security and strengthen efforts to combat terrorism, making air travel safer and more secure for all passengers.

Multi-scale multimodal deep learning framework for Alzheimer's disease diagnosis

Multimodal neuroimaging data, including magnetic resonance imaging (MRI) and positron emission tomography (PET), provides complementary information about the brain that can aid in Alzheimer's disease (AD) diagnosis. However, most existing deep learning methods still rely on patch-based extraction from neuroimaging data, which typically yields suboptimal performance due to its isolation from the subsequent network and does not effectively capture the varying scales of structural changes in the cerebrum. Moreover, these methods often simply concatenate multimodal data, ignoring the interactions between them that can highlight discriminative regions and thereby improve the diagnosis of AD. To tackle these issues, we develop a multimodal and multi-scale deep learning model that effectively leverages the interaction between the multimodal and multiscale of the neuroimaging data. First, we employ a convolutional neural network to embed each scale of the multimodal images. Second, we propose multimodal scale fusion mechanisms that utilize both multi-head self-attention and multi-head cross-attention, which capture global relations among the embedded features and weigh each modality's contribution to another, and hence enhancing feature extraction and interaction between each scale of MRI and PET images. Third, we introduce a cross-modality fusion module that includes a multi-head cross-attention to fuse MRI and PET data at different scales and promote global features from the previous attention layers. Finally, all the features from every scale are fused to discriminate between the different stages of AD. We evaluated our proposed method on the ADNI dataset, and the results show that our model achieves better performance than the state-of-the-art methods.

Predicting autism spectrum disorder through sentiment analysis with attention mechanisms: a deep learning approach

Autism spectrum disorder (ASD) is considered a spectrum disorder. The availability of technology to identify the characteristics of ASD will have major implications for clinicians. In this article, we present a new autism diagnosis method based on attention mechanisms for behavior modeling-based feature embedding along with aspect-based analysis for a better classification of ASD. The hybrid model comprises a convolutional neural network (CNN) architecture that integrates two bidirectional long short-term memory (BiLSTM) blocks, together with additional propagation techniques, for the purpose of classification the origins of Autism Tweet dataset; the proposed work takes Autism Tweet dataset and preprocesses them to employ n-gram to extract features of which the features of the ASD behavior are fed to generate the significant behavior for classification. The model takes into account both behavior-guided features across every aspect of the Class/ASD to provide higher accuracy using Adam optimizer. The experimental values inferred that the n-BiLSTM technique reaches maximum accuracy with 98%.

Uncertainty modeling for inductive knowledge graph embedding.

In the process of refining Knowledge Graphs (KGs), new entities emerge, and old entities evolve, which usually updates their attribute information and neighborhood structures. This results in a distribution shift problem for entity features in the embedding space during graph representation learning. Most of existing inductive knowledge graph embedding methods focus mainly on the representation learning of new entities, neglecting the negative impact caused by distribution shift of entity features. In this paper, we use the skill of mean and variance reconstruction to develop a novel inductive knowledge graph embedding model named EDSU for processing the shift of entity feature distribution. Specifically, by assuming that the embedding feature of entity follows multivariate Gaussian distribution, the reconstruction combines the distribution characteristics of components in an entity embedding vector with neighborhood structure information of a set of entity embedding vectors, in order to alleviate the deviation of data information between intra-entity and inter-entity. Furthermore, the connection between the entity features distributions before and after the shift is established, which guides the model training process and provides an interpretation on the rationality of such handling distribution shift in view of distributional data augmentation. Extensive experiments have been conducted and the results demonstrate that our EDSU model outperforms previous state-of-the-art baseline models on inductive link prediction tasks.

A deep-learning system integrating electrocardiograms and laboratory indicators for diagnosing acute aortic dissection and acute myocardial infarction.

Acute Stanford Type A aortic dissection (AAD-type A) and acute myocardial infarction (AMI) present with similar symptoms but require distinct treatments. Efficient differentiation is critical due to limited access to radiological equipment in many primary healthcare. This study develops a multimodal deep learning model integrating electrocardiogram (ECG) signals and laboratory indicators to enhance diagnostic accuracy for AAD-type A and AMI. We gathered ECG and laboratory data from 136 AAD-type A and 141 AMI patients at Zigong Fourth People's Hospital (January 2019 to December 2023) for training and validation. Utilizing ResNet-34 (residual network), we extracted ECG features and combined them with laboratory and demographic data. We assessed logistic regression, RandomForest, XGBoost, and LightGBM models, employing shapley additive explanations (SHAP) for feature importance analysis. Data from 30 AMI and 32 AAD-type A patients (January to September 2024) were used as a prospective test set. Incorporating ECG features significantly improved model's AUC value, with the RandomForest achieving the best performance (AUC 0.98 on validation, 0.969 on test). SHAP analysis revealed that troponin and D-dimer, along with the embedding features of ECG extracted by the deep neural network, are key characteristics for differentiating AAD-type A and AMI. ECG features are valuable for distinguishing AAD-type A and AMI, offering a novel tool for rapid cardiovascular disease diagnosis through multimodal data fusion and deep learning.

A lightweight and rapidly converging transformer based on separable linear self-attention for fault diagnosis

Abstract Reaching reliable decisions on equipment maintenance is facilitated by the implementation of intelligent fault diagnosis techniques for rotating machineries. Recently, the Transformer model has demonstrated exceptional capabilities in global feature modeling for fault diagnosis tasks, garnering significant attention from the academic community. However, it lacks sufficient prior knowledge regarding rotation invariance, scale, and shift, necessitating pre-training on extensive datasets. In comparison, contemporary convolutional neural networks exhibit greater ease of optimization. This limitation becomes particularly evident when applying the Transformer model in fault diagnosis scenarios with limited data availability. Moreover, the increasing the number of parameters and FLOPs. Pose a challenge to its suitability for mobile services due to the limited computational resources available on edge devices. To mitigate these issues, this paper introduces a novel lightweight Transformer (SepFormer) based on separable linear self-attention (LSA) for fault diagnosis task. The SepFormer performs a novel sequence-level feature embedding to better leverage the inductive bias inherent in the convolutional layers. Furthermore, it integrate a novel separable LSA mechanism into the Transformer architecture, effectively mitigating the computational burden concerns and significantly enhancing the training convergence speed. Extensive experiments are conducted extensively on a bearing fault dataset and gear fault dataset. The experimental results demonstrate that the SepFormer achieves a top-1 accuracy exceeding state-of-the-art approaches by more than 5%, while utilizing the fewest FLOPs. Moreover, the optimizability of SepFormer surpasses that of CNN, ensuring its superior preservation of inductive bias.

Harmonized system code classification using supervised contrastive learning with sentence BERT and multiple negative ranking loss

PurposeInternational trade transactions, extracted from customs declarations, include several fields, among which the product description and the product category are the most important. The product category, also referred to as the Harmonised System Code (HS code), serves as a pivotal component for determining tax rates and administrative purposes. A predictive tool designed for product categories or HS codes becomes an important resource aiding traders in their decision to choose a suitable code. This tool is instrumental in preventing misclassification arising from the ambiguities present in product nomenclature, thus mitigating the challenges associated with code interpretation. Moreover, deploying this tool would streamline the validation process for government officers dealing with extensive transactions, optimising their workload and enhancing tax revenue collection within this domain.Design/methodology/approachThis study introduces a methodology focused on the generation of sentence embeddings for trade transactions, employing Sentence BERT (SBERT) framework in conjunction with the Multiple Negative Ranking (MNR) Loss function following a contrastive learning paradigm. The procedure involves the construction of pairwise samples, including anchors and positive transactions. The proposed method is evaluated using two publicly available real-world datasets, specifically the India Import 2016 and United States Import 2018 datasets, to fine-tune the SBERT model. Several configurations involving pooling strategies, loss functions, and training parameters are explored within the experimental setup. The acquired representations serve as inputs for traditional machine learning algorithms employed in predicting the product categories within trade transactions.FindingsEncoding trade transactions utilising SBERT with MNR loss facilitates the creation of enhanced embeddings that exhibit improved representational capacity. These fixed-length embeddings serve as adaptable inputs for training machine learning models, including support vector machine (SVM) and random forest, intended for downstream tasks of HS code classification. Empirical evidence supports the superior performance of our proposed approach compared to fine-tuning transformer-based models in the domain of trade transaction classification.Originality/valueOur approach generates more representative sentence embeddings by creating the network architectures from scratch with the SBERT framework. Instead of exploiting a data augmentation method generally used in contrastive learning for measuring the similarity between the samples, we arranged positive samples following a supervised paradigm and determined loss through distance learning metrics. This process involves continuous updating of the Siamese or bi-encoder network to produce embeddings derived from commodity transactions. This strategy aims to ensure that similar concepts of transactions within the same class converge closer within the feature embedding space, thereby improving the performance of downstream tasks.

A hybrid Hadoop-based sentiment analysis classifier for tweets associated with COVID-19 utilizing two machine learning algorithms: CNN, and fuzzy C4.5

In recent years, research on opinion mining from X (formerly Twitter) has rapidly advanced, focusing on processing tweets to determine user sentiments about events. Many researchers prefer using machine and deep learning techniques for this analysis. This work proposes a novel approach integrating the C4.5 procedure, fuzzy rule patterns, and convolutional neural networks. The approach involves six steps: pre-processing to remove noisy data, vectorizing tweets with word embedding, extracting sentiment and contextual features using convolutional neural networks, fuzzifying outputs with a Gaussian fuzzifier to handle ambiguity, constructing a fuzzy tree and rule base using a fuzzy version of C4.5, and classifying tweets with fuzzy General Reasoning. This method combines the benefits of convolutional neural networks and C4.5 while addressing imprecise data with fuzzy logic. Implemented on a Hadoop framework-based cluster with five computing units, the approach was extensively tested. The results showed that the model performs exceptionally well on the COVID-19_Sentiments dataset, surpassing other classification algorithms with a precision rate of 94.56%, false-negative rate of 5.28%, classification rate of 95.15%, F1-score of 94.63%, kappa statistic of 95.12%, execution time of 11.81 s, false-positive rate of 4.26%, error rate of 4.26%, specificity of 95.74%, recall of 94.72%, stability with a mean deviation standard of 0.09%, convergence starting around the 75th round, and significantly reduced complexity in terms of time and space.