Understanding customer satisfaction in e-commerce is crucial for businesses to remain competitive. While traditional feedback analysis methods are labour-intensive and subjective, machine learning advances have enabled more efficient and scalable sentiment analysis. However, existing models struggle with aspect-based sentiment analysis (ABSA), particularly in detecting implicit aspects and handling mixed sentiments. This paper presents a multi-model machine learning pipeline designed to enhance ABSA by integrating fine-tuned Large Language Models (LLMs) with BERT and RoBERTa-based models. The pipeline consists of an LLM-generated synthesized annotated feedback model, a BERT-based aspect detection model, a RoBERTa-based ABSA model, and an LLM-based ABSA model for handling implicit aspects and mixed sentiments. Additionally, a RoBERTa-based model is employed for overall sentiment detection. By leveraging both manually annotated and synthetic data, the pipeline improves sentiment classification accuracy and aspect coverage, even in data-scarce environments. The results demonstrate that combining multiple models enhances detection accuracy compared to single-model approaches. This study provides a scalable and effective solution for e-commerce feedback analysis, offering businesses valuable insights for improving customer experience and decision-making.

Chronic diseases for children pose serious challenges from a health management perspective. When not implemented in a well-designed manner, an inefficient management platform can have a significant negative impact on patients and the utilization of health care resources. Innovations of recent years in information technology, artificial intelligence and machine learning provide possibilities to design and implement knowledge-based systems and platforms that follow-up, monitor and advise child patients with a chronic disease in an automated manner. In this article we propose the Artificial Intelligence Chronic Management System that combines artificial intelligence, knowledge graph, big data and internet of things in a platform to offer an optimized solution from the perspective of treatment and utilization of resources. The system includes patient and hospital clients, data storage and analytic tools for decision support relying on AI-based services. We illustrate the functionality of the system through different situations frequently occurring in pediatric wards. To assess the feasibility of the AI component, we utilize real life health care data from a hospital in China to develop a classification model for patients with asthma. To provide a more qualitative assessment at the same time, we discuss how the Artificial Intelligence Chronic Management System conforms to the requirements set forth by the standard Chronic Care Model.

Classification learning is a very complex process whose success and failure ratio depends on a high amount of elements. One of them is the representation mean used for the data that is employed in the process. Granularity of the data used for classification learning purposes can affect dramatically the success and failure ratio of the obtained classification. In this paper, multi-granular fuzzy linguistic modelling methods are applied over the classification learning data in order to modify their granularity and increase the classification success ratio. Thanks to multi-granular fuzzy linguistic modelling methods, it is possible to automatically modify the data granularity in order to determine which data representation is the one that provides the better classification results in the learning process.

Obtaining good classification results using supervised learning methods is critical if we want to obtain a high level of precision in the classification processes. The training data used for the learning process play a very important role in achieving this objective. Therefore, it is important to represent the data in a way that best expresses its meaning. For this purpose, we propose to apply linguistic modeling methods in order to obtain a linguistic representation. With the help of multigranular linguistic modeling, data can be transformed and expressed using different (unbalanced) linguistic label sets. Expressing the data using linguistic expressions instead of numbers increases the readability and reduces the complexity of the problem, and data recovering methods allow us to manually control the level of precision. In this paper, several datasets are transformed and utilized for classification tasks using several supervised learning algorithms. For each combination of datasets and algorithms, the data have been expressed using several linguistic label sets that have different granularity values. After carrying out the testing processes, we can conclude that, in some cases, reducing data complexity leads to better classification results. Therefore, it is found that linguistic representation of the training data with just the necessary and sufficient precision can improve the reliability of the classification process.

Representing different types of uncertainty present in data is an important problem in developing machine learning algorithms. In this paper we focus on objective function-based fuzzy clustering methods, and propose an interval-valued extension of an approach based on a general objective function. In the model, both membership and possibilistic typicality values are incorporated to overcome various problems of previous clustering approaches. To illustrate the usefulness of the proposals, we perform numerical experiments on eight different fuzzy-possibilistic clustering methods and three data-sets to evaluate the performance in a binary classification problem. We find that the interval-valued extension offers improved performance compared to various approaches from the possibilistic clustering literature.

Process interruptions in (very) large production systems are difficult to deal with. Modern processes are highly automated; data is collected with sensor technology that forms a big data context and offers challenges to identify coming failures from the very large sets of data. Feature selection is intended to reduce the complexity of identifying cases with high possibility of failure by excluding numerous factors in the process systems. We use fuzzy entropy as the basis of a feature selection method and we show how the outcome of feature selection can be utilized to further failure prediction steps.

Knowledge mobilization is a transition from the prevailing knowledge management to a new methodology through some innovative methods for knowledge representation, formation, and development and for knowledge retrieval and distribution. The context is industrial processes and finding solutions to complex problems that arise and for which at least partial solutions have been documented. The fact that a problem has been solved before normally makes it easier to solve it again and the existence of documents that describe how it was solved supports the problem-solving process. But documents that describe the problem solving have to be retrieved from a large database of documents and the information that describes the content of a document is not precise. We show that fuzzy ontology will be useful for finding a sufficiently small set of documents that are relevant for the problem solving even if they are imprecisely classified with keywords.

Knowledge mobilisation is a transition from the prevailing knowledge management technology to a new methodology and some innovative methods for knowledge representation, formation and development and for knowledge retrieval and distribution. We show that fuzzy ontology will be useful to represent real world knowledge and that approximate reasoning schemes can give us answers which are sufficiently good for real world situations in which we need sufficiently good knowledge. We demonstrate the knowledge mobilisation approach by showing how amateurs can become wine connoisseurs with support from the technology.