Jul 2026· Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology· 0 citations· 25 references
TL;DR
Experimental results show that BERT-LSTM's superior performance stems from its ability to capture bidirectional contextual dependencies, improving contextual modelling of complex linguistic constructs, however, sarcasm and negation remain important sources of error.
Abstract
Sentiment analysis is a crucial problem in natural language processing (NLP), allowing algorithms to identify and analyze subjective information in text. Despite the rapid evolution of sentiment analysis techniques, prior research has largely lacked unified, reproducible evaluations directly comparing conventional machine learning models with deep learning architectures, particularly across diverse word embedding strategies. This study presents a comprehensive comparative analysis of sentiment classification models, systematically evaluating both conventional machine learning techniques (e.g., SVM, Naive Bayes, Decision Tree, Random Forest) and advanced deep learning architectures (CNN, RNN, LSTM, BERT-LSTM) across various datasets (IMDB, Yelp, Amazon). We employ various word embedding methodologies, including TF-IDF, Word2Vec, GloVe, FastText, and BERT, to assess their impact on model efficacy. BERT-LSTM achieved the best overall performance across datasets (IMDB: 94.0%, Yelp: 94.0%, Amazon: 96.0% accuracy). In contrast, traditional machine learning models such as Naive Bayes and SVM generally achieved lower accuracy than deep contextual models across the evaluated datasets. Experimental results further show that BERT-LSTM's superior performance stems from its ability to capture bidirectional contextual dependencies, improving contextual modelling of complex linguistic constructs. However, sarcasm and negation remain important sources of error. To enhance repeatability, we provide detailed hyperparameter tuning procedures, hardware configurations, and all preprocessing steps, model configurations, and evaluation protocols are described in the manuscript to support independent replication. A targeted error analysis is conducted to systematically examine misclassifications arising from ambiguous language use and sarcastic expressions, highlighting key limitations of existing models. In light of BERT's computational requirements, we investigate pragmatic deployment techniques, such as quantisation (INT8), ONNX Runtime optimisation, and lightweight transformer alternatives (DistilBERT, TinyBERT), that strike a balance between performance and efficiency. INT8 quantisation reduced inference latency by approximately 35–40% under the reported experimental configuration, accompanied by a marginal 1.8% decrease in accuracy. DistilBERT achieved comparable accuracy (91.5%) while processing significantly faster. Although the current experiments focus on English-language datasets, the proposed framework is designed to support future cross-lingual extensions using multilingual models such as mBERT, XLM-R, and IndicBERT. Our findings offer a cohesive perspective on optimizing sentiment models, integrating theoretical frameworks with practical implementation issues, and facilitating the development of scalable, efficient, and interpretable sentiment analysis systems.
— By comparing and analysing the performance of several machine learning algorithms on fine-grained sentiment classification problems to examine their suitability and shortcomings for use as models in sentiment analysis. Sentiment analysis remains a prominent research area in Natural Language Processing (NLP). However, systematic comparisons of whether these methods demonstrate superiority in fine-grained classification settings have yet to be conducted. Based on the GoEmotions dataset that includes 28 sentiment classes as our experiment’s reference for evaluating four kinds of models: logistic regression, BiLSTM, BERT, and the large-scale language model Claude (claude-haiku-4-5). The experiments employ TF-IDF feature extraction, sequence encoding, pre-training with fine-tuning, and zero-shot and few-shot prompting strategies. Based on this experiment, the model’s generalisation capabilities improve as architectural depth increases. The accuracy of logistic regression was only 55.11%; BERT obtained the highest F1 score and performed the best overall, and bidirectional pre-trained representations were most valuable among other methods. However, large language models perform significantly worse on the 28-class classification task in zero-shot settings, suggesting that they are better suited for generative and open-ended emotional interaction than for standardized classification benchmarks.
Shangjiafeng Guo· International journal of eng...· 0 citations
With the rapid growth of e-commerce platforms, understanding customer feedback has become crucial for businesses to improve their products and services. Sentiment analysis, a subfield of Natural Language Processing (NLP), plays a vital role in extracting opinions from user-generated content such as product reviews. This project aims to perform sentiment analysis on e-commerce product reviews using a hybrid deep learning approach. A CNN-LSTM (Convolutional Neural Network - Long Short-Term Memory) model was implemented to classify customer reviews into positive and negative sentiments. The CNN layer effectively captures local features from the text, while the LSTM layer processes sequential dependencies for better contextual understanding. The dataset was preprocessed using techniques such as tokenization, stop word removal, and padding, followed by word embedding for semantic representation. Experimental results demonstrate that the CNN-LSTM model achieves an impressive accuracy of 95%, outperforming traditional machine learning models. The proposed model provides an efficient solution for automating sentiment classification in large-scale e-commerce platforms.
Mounika Garikapati, Nageswara Rao Kapu· International Journal For Mu...· 0 citations
Understanding emotional tone and sentiment in text has driven significant advancements in Natural Language Processing (NLP), particularly in sentiment analysis and emotion detection. This study addresses the challenge of developing effective NLP tools for low-resource languages, focusing on the Hausa language. By leveraging ensemble methods and pre-trained transformer models like BERT and XLM-R, along with traditional classifiers such as Logistic Regression, SVM, Naive Bayes, Random Forest, and XGBoost, we aim to improve sentiment analysis and emotion detection for Hausa text. Utilizing a balanced sentiment dataset (9,958 samples) and a complex multi-label emotion dataset (19,757 samples across 11 categories), we benchmark individual classifiers, voting ensembles, and deep contextual models. For sentiment analysis, a Hard Voting Ensemble of TF-IDF-vectorized base learners achieved a highly competitive F1-score of 0.8748. However, Transformer models significantly outperformed traditional baselines, with Multilingual BERT (mBERT) achieving a peak F1-score of 0.8983. In the multi-label emotion detection task, individual traditional models struggled with label sparsity, yielding low Subset Accuracy scores (2.88% to 8.30%) and moderate Micro-F1 scores. Standard Hard Voting ensembles further underperformed due to discrete prediction conflicts. To resolve this, a Probability-based Majority Voting mechanism with calibrated thresholding (0.3) was introduced, boosting the Micro-F1 to 0.3825 and reducing the Hamming Loss to 0.1967. Ultimately, XLM-RoBERTa emerged as the superior architecture, achieving a Subset Accuracy of 0.1545, a Micro-F1 of 0.4275, and the lowest Hamming Loss of 0.1804. This research establishes a rigorous benchmark for Hausa NLP, highlighting the indispensable role of subword tokenization, contextual embeddings, and threshold calibration in handling the morphological richness and multi-label complexities of low-resource languages
Taiwo Kolajo, Kabir Garba· International journal of res...· 0 citations
Sentiment analysis is widely used to extract opinions from textual data; however, its application to morphologically rich languages such as Turkish remains challenging. This study investigates the optimization of classical machine learning classifiers and ensemble learning strategies for binary Turkish sentiment analysis under a unified experimental framework. Several ML models are trained on a balanced dataset of user reviews, including Linear Support Vector Machine Classifier (LSVMC), Multinomial Naïve Bayes (MNB), and Logistic Regression (LR). Their outputs were further combined using Ensemble Learning (EL) models, namely Majority Voting (MVEL) and Stacking (SEL). Results demonstrate that the SEL Classifier outperforms all examined models, achieving 92.80% accuracy at the cost of increased computational complexity. Among the examined individual models, LSVMC (92%), MNB (92%), and LR (92%) had the best accuracy. While the study does not aim to achieve state-of-the-art performance with deep or transformer-based architectures, the results demonstrate that optimized classical models remain highly effective in Turkish SA, and their accuracy can be further improved with EL mechanisms.
Ahmad Bwidani, Ali Karah Bash· Uludağ University Journal of...· 0 citations
The work provides a reproducible, explainable, operationally applicable model of sentiment analysis in operationally sensitive, high-stakes Twitter sentiment analysis, and validate the hypothesis that hybrid stacking is an effective method for leveraging the complementary nature of lexical and contextual representations.
D. Abate, Nilay Mistry· International Research Journ...· 0 citations
Sentiment analysis has become an essential Natural Language Processing (NLP) technique for extracting opinions and emotions from textual data generated through social media, online reviews, blogs, and customer feedback. Although deep learning models such as Long Short-Term Memory (LSTM), Bidirectional Encoder Representations from Transformers (BERT), and other transformer-based architectures have achieved remarkable accuracy in sentiment classification, their black-box nature limits interpretability and user trust. Explainable Artificial Intelligence (XAI) addresses this limitation by providing transparent and understandable explanations for model predictions, enabling users to identify the key words, phrases, and contextual features that influence sentiment classification. This paper presents a comprehensive study of Explainable AI techniques applied to sentiment analysis, focusing on both model-agnostic methods, including Local Interpretable Model-Agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP), and attention-based explanation mechanisms. The proposed framework integrates text preprocessing, feature extraction using TF-IDF and contextual embeddings, sentiment classification through machine learning and deep learning models, and explanation generation to improve model transparency. Performance is evaluated using publicly available sentiment datasets based on accuracy, precision, recall, F1-score, and explanation quality. Experimental results demonstrate that XAI techniques significantly enhance the interpretability of sentiment prediction without substantially compromising classification performance. Furthermore, explainable sentiment analysis supports fairness assessment, bias detection, regulatory compliance, and informed decision-making in critical domains such as healthcare, finance, education, and social media analytics. The findings highlight that integrating explainability with sentiment analysis not only increases model reliability but also promotes greater user confidence and responsible deployment of artificial intelligence systems.
Aishwarya P. A., N. K· International Journal of Res...· 0 citations