Search

The rapid growth of Internet-enabled applications, such as social media platforms, e-commerce sites, and blogs, has led to a surge in user-generated content. This vast amount of data has made sentiment analysis increasingly valuable. Modern Aspect-Based Sentiment Analysis (ABSA) offers a more detailed approach by identifying sentiment trends related to specific aspects within the text. However, the challenge lies in analyzing reviews that are often short, unstructured, and filled with slang and emotive language, making it difficult to gauge customer opinions accurately. To address these issues, we proposed an effective hybrid approach “RoBERTa-1D-CNN-BiLSTM” for ABSA. Initially, the pre-trained Robustly Optimized BERT approach (RoBERTa) and One Dimensional Convolutional Neural Network (1D-CNN) models are used to extract features at the aspect level from the context of the review, following which classification is performed using Bidirectional Long Short-Term Memory (BiLSTM). The approach is evaluated on three cross-domain standards datasets, yielding an accuracy of 92.33%. The results of the experiments show that it surpasses the current leading methods in sentiment analysis and product recommendation.

As technology advances, most people use social media sites like Twitter, Facebook, and Flickr to share information and communicate with others. The volume of free-text data is growing daily due to the widespread use of these social media platforms. These platforms contain a substantial amount of unstructured information. Patient opinions expressed on social media platforms play a significant role in healthcare improvement and impact health-related policymaking. In this research, we introduce a machine learning approach for the optimal identification of healthcare-related features. This approach is based on a novel synthetic method. Additionally, we employ an entropy-based technique to classify free-text comments from hospital data into positive, negative or neutral. The experimental results and evaluations show 85%, 82.3%, 78.2% and 87% accuracy between ratings of health care. We observed that there is a minor association between our technique, expert opinion and patient interviews. Through the use of machine learning techniques, we achieve an accuracy level that suggests we are capable of providing an accurate and reasonable assessment of the ideal healthcare center for a patient. Our proposed novel framework predicts the healthcare experience at hospitals based on patient reviews posted on social media. This innovative approach outperforms traditional methods, such as surveys and expert opinions.

Serious consequences including septicemia and amputations can result from complex wounds, which is a serious healthcare concern. In addition, there are currently only a few choices for management, which justifies the search for novel, highly effective wound-healing medications. This research work was aimed at fabricating chitosan-based Alkanna tinctoria and Mupirocin nanoparticles by ionic gelation technique for burn wound management. Preliminary studies were conducted, and the prepared nanoparticles were characterized by various techniques that involve, high-performance liquid chromatography for the detection of components in A. tinctoria root extract, ATR-FTIR, particle size, zeta potential, percent drug content (DC%), percent entrapment efficiency (EE%), scanning electron microscopy, and transmission electron microscopy (TEM) for surface morphology. The optimized formulation CS-AT-MU-NPs3 shows a particle size of 340.8 ± 34.46 nm and positive zeta potential 27.3 ± 3.10 mV. In vitro drug release study was also performed, which demonstrated improved and controlled release of the drug from the nanoparticles. The CS-AT-MU-NPs3 exhibited a maximum release up to 92.61% (AT) and 88.35% (MU). Antibacterial and antifungal activities of the formulation were also accessed by utilizing the agar well diffusion technique. The combination of AT and MU in chitosan-based nanoparticles was significantly effective against bacterial and fungal strains like Staphylococcus aureus, Pseudomonas aeruginosa, and Candida albicans as compared to other formulations. The skin irritation study was also conducted, which shows that the prepared formulation did not cause any observable changes to the skin in terms of inflammation, erythema, edema, or any other symptoms associated with skin irritation. All the chitosan-based nanoparticles showed almost 75% reduction in wound contraction, while the optimized formulation CS-AT-MU-NPs3 showed complete wound healing on the 15th day. Based on the results, it can be assumed that chitosan-based nanoparticles containing A. tinctoria and Mupirocin demonstrated good wound healing and could be used to effectively manage burn wounds of any description.

With the increase in Technology Enhanced Learning (TEL), the effective retrieval and availability of Learning Objects (LOs) for course designers is a significant concern. Text-based LOs can be accessed from structured LO repositories (LORs) and unstructured sources. Different LOR structures and semantically diversified LOs hinder the process of course designing because of various repository formats and standards. The proposed research provides an innovative Machine Learning and filter based context-aware LO recommender System for the designing of course. The proposed model helps to easily search and access diversified and semantically related LOs. The proposed solution will also help the academia to design customized courses gratifying higher education curriculum needs while reusing existing LOs. Through automatic feature extraction and classification, the proposed system -- Dynamic Recommendation of Filtered LOs (DRFLO) -- retrieves a set of semantically relevant LOs from heterogeneous LORs. Using the additional recommendation and ranking models, LOs are contextually ranked. Finally, a ranked list of LOs is recommended on a simple query from the course designer according to learning preferences while designing a course. The performance of the proposed model is quite remarkable in terms of precision, recall and F-measure. Furthermore, the experimental results have also been compared with a common search engine and an accuracy of 93% was achieved. The proposed research further produced, pre-test and post-test experiments to ensure the validity of the DRFLO approach. The proposed DRFLO System is a machine learning based research effort which provides an efficient mean for retrieving LOs while designing a course.

Controlled-release microparticles offer a promising avenue for enhancing patient compliance and minimizing dosage frequency. In this study, we aimed to design controlled-release microparticles of Glipizide utilizing Eudragit S100 and Methocel K 100 M polymers as controlling agents. The microparticles were fabricated through a simple solvent evaporation method, employing various drug-to-polymer ratios to formulate different controlled-release batches labeled as F1 to F5. Evaluation of the microparticles encompassed a range of parameters including flow properties, particle size, morphology, percentage yield, entrapment efficiencies, percent drug loading, and dissolution studies. Additionally, various kinetic models were employed to elucidate the drug release mechanism. Furthermore, difference and similarity factors were utilized to compare the dissolution profiles of the tested formulations with a reference formulation. The compressibility index and angle of repose indicated favorable flow properties of the prepared microparticles, with values falling within the range of 8 to 10 and 25 to 29, respectively. The particle size distribution of the microparticles ranged from 95.3 to 126 μm. Encouragingly, the microparticles exhibited high percent yield (ranging from 66 to 77%), entrapment efficiency (80 to 96%), and percent drug loading (46 to 54%). All formulated batches demonstrated controlled drug release profiles extending up to 12 hours, with glipizide release following an anomalous non-Fickian diffusion pattern. However, the drug release profiles of the reference formulation and various polymeric microparticles did not meet the acceptable limits of difference and similarity factors. In-vivo studies revealed sustained hypoglycemic effects over a 12-hour period, indicating the efficacy of the controlled-release microparticles. Overall, our findings suggest the successful utilization of polymeric materials in designing controlled-release microparticles, thereby reducing dosage frequency and potentially improving patient compliance. [ABSTRACT FROM AUTHOR]

Crime hotspot detection is essential for law enforcement agencies to allocate resources effectively, predict potential criminal activities, and ensure public safety. Traditional methods of crime analysis often rely on manual, time-consuming processes that may overlook intricate patterns and correlations within the data. While some existing machine learning models have improved the efficiency and accuracy of crime prediction, they often face limitations such as overfitting, imbalanced datasets, and inadequate handling of spatiotemporal dynamics. This research proposes an advanced machine learning framework, CHART (Crime Hotspot Analysis and Real-time Tracking), designed to overcome these challenges. The proposed methodology begins with comprehensive data collection from the police database. The dataset includes detailed attributes such as crime type, location, time and demographic information. The key steps in the proposed framework include: Data Preprocessing, Feature Engineering that leveraging domain-specific knowledge to extract and transform relevant features. Heat Map Generation that employs Kernel Density Estimation (KDE) to create visual representations of crime density, highlighting hotspots through smooth data point distributions and Hotspot Detection based on Random Forest-based to predict crime likelihood in various areas. The Experimental evaluation demonstrated that CHART shows superior performance over benchmark methods, significantly improving crime detection accuracy by getting 95.24% for crime detection-I (CD-I), 96.12% for crime detection-II (CD-II) and 94.68% for crime detection-III (CD-III), respectively. By designing the application with integrating sophisticated preprocessing techniques, balanced data representation, and advanced feature engineering, the proposed model provides a reliable and practical tool for real-world crime analysis. Visualization of crime hotspots enables law enforcement agencies to strategize effectively, focusing resources on high-risk areas and thereby enhancing overall crime prevention and response efforts. [ABSTRACT FROM AUTHOR]

This study aimed to examine the anti-diabetic effects of an unexplored medical plant Sorbaria tomentosa Lindl. Rehder using in vitro and in vivo approaches. The extracts were tested as inhibitors of α-glucosidase and α-amylase following standard protocols. Methanolic extract was analyzed via gas chromatography-mass spectrometry (GC-MS) for tentative identification of the secondary metabolites. GC-MS analysis revealed the presence of several compounds. α-Amylase was more potently inhibited by chloroform and methanolic extracts (27 and 40 µg mL−1, respectively), whereas α-glucosidase was more potently inhibited by methanolic extract (IC50 = 530 µg mL−1). Methanolic extract was also subjected to in vivo studies using an alloxan-induced diabetes rat model. Diabetic animals treated with 150 mg kg−1 body weight dose of methanolic extract cause a steady decrease in blood glucose levels (529.16, 446.66, 348.00, 269.33, and 165.5 mg dL−1, respectively, on days 1, 7, 14, 21, and 28). At 300 mg kg−1 dose, the blood glucose level was decreased to 111.83 mg dL−1 on day 28. Blood biochemistry results indicated that treatment with methanolic extract normalized the elevated parameters including cholesterol, triglycerides, low-density lipoprotein, serum creatinine, blood urea, uric acid, serum glutamate pyruvate transaminase, bilirubin, alkaline phosphatase, and aspartate aminotransferase in diabetic animals.

Microblogs and social media sites have gained a central place and people use these platforms to express their opinions, sentiments, and thoughts about products, news, events, blogs, etc. Sentiment analysis is the process of exploring opinions and sentiments in user reviews and tweets. This area is still in its early developmental phase and requires imperative improvements on various issues. One of the main issues is multilingual tweets and reviews. Earlier sentiment analysis techniques only classified the text of a specific language, i.e., English, Turkish, etc. The accuracy of these techniques decreases in the presence of multilingual text. Existing methods are domain oriented. Using BERT and a lexicon, we propose a method for sorting out multilingual text and improving the polarity calculation of reviews. Experimental results reveal that our proposed technique achieved 90.14% accuracy and outperformed existing aspect-based sentiment analysis techniques. [ABSTRACT FROM AUTHOR]

Heart disease remains a leading cause of morbidity and mortality worldwide, highlighting the need for improved diagnostic methods. Traditional diagnostics face limitations such as reliance on single-modality data and vulnerability to apparatus faults, which can reduce accuracy, especially with poor-quality images. Additionally, these methods often require significant time and expertise, making them less accessible in resource-limited settings. Emerging technologies like artificial intelligence and machine learning offer promising solutions by integrating multi-modality data and enhancing diagnostic precision, ultimately improving patient outcomes and reducing healthcare costs. This study introduces Heart-Net, a multi-modal deep learning framework designed to enhance heart disease diagnosis by integrating data from Cardiac Magnetic Resonance Imaging (MRI) and Electrocardiogram (ECG). Heart-Net uses a 3D U-Net for MRI analysis and a Temporal Convolutional Graph Neural Network (TCGN) for ECG feature extraction, combining these through an attention mechanism to emphasize relevant features. Classification is performed using Optimized TCGN. This approach improves early detection, reduces diagnostic errors, and supports personalized risk assessments and continuous health monitoring. The proposed approach results show that Heart-Net significantly outperforms traditional single-modality models, achieving accuracies of 92.56% for Heartnet Dataset I (HNET-DSI), 93.45% for Heartnet Dataset II (HNET-DSII), and 91.89% for Heartnet Dataset III (HNET-DSIII), mitigating the impact of apparatus faults and image quality issues. These findings underscore the potential of Heart-Net to revolutionize heart disease diagnostics and improve clinical outcomes. [ABSTRACT FROM AUTHOR]