1 June 2024 in Indonesian Journal of Electrical Engineering and Computer Science
Social media sentiment classification was an essential consideration in natural language processing (NLP) for evaluating normal people’s perspectives on a given topic. With Twitter’s massive rise in popularity in recent years, the capacity to extract information about public sentiment from tweets became a major focus. This paper not only analyzed public sentiment through data from Twitter but introduced a novel ensemble approach in the methods employed for Twitter sentiment classification. Real-time tweets on various topics, including “covid,” “crime,” “spam,” “flipkart,” “migraine,” and “airlines,” were extracted and thoroughly examined to gain insight into public opinions. Leveraging the Twitter API for real-time tweet extraction, natural language processing techniques were applied to clean the tweet data. Subsequently, we applied several machine learning (ML) algorithms Naïve Bayes, decision tree (DT), random forest (RF), logistic regression (LGR), and deep learning (DL) algorithms recurrent neural network (RNN), LSTM, and GRU individually. Later, we proposed a novel ensemble of ML and DL algorithms for sentiment classification, with a novel emphasis on ensemble techniques and enhanced the accuracy with a significance compared to individual ML or DL model applied. The experimental results demonstrated that our novel ensemble approach achieved high accuracy when compared to existing work.
14 July 2025 in International Research Journal of Multidisciplinary Technovation
Accurate classification of medicinal plant images into high-level categories and specific sub-groups is essential for various applications, including agriculture, plant research, and conservation. This paper proposes a multi-stage deep learning approach to enhance the precision of medicinal plant image classification. In the first stage, known as Broad Classification, CNN and pre-trained models such as VGG16, ResNet50 and EfficientNetB0 are utilized to categorize images into high-level groups, including "Medicinal Plants," "Fruit-Related Plants," and "Flower-Related Plants." The model is fine-tuned using data augmentation techniques to ensure robust learning and generalization. In the second stage, referred to as Detailed Classification, separate models are trained for each high-level group to classify images into specific sub-groups within that category. The architecture of these models is adjusted to accommodate the unique number of classes in each sub-group. Each model undergoes training with optimized hyperparameters and is evaluated based on precision, recall, F1-score, and accuracy. The proposed multi-stage method demonstrates the ability to handle both broad and fine-grained medicinal plant classifications effectively, showcasing an improvement in classification performance over traditional single-stage models. This approach highlights the potential for deep learning to contribute to more precise and practical medicinal plant image classification solutions.