Search

Machine learning revolutionizes accuracy in diverse fields such as disease diagnosis, speech understanding, and sentiment analysis. However, its intricate architecture often obscures the decision-making process, creating a "black box" that hinders trust and limits its potential. This lack of transparency poses significant challenges, particularly in critical fields like the healthcare system. We present OntoXAI, a Semantic Web Rule Language (SWRL) based Explainable Artificial Intelligence (XAI) approach to address these challenges. OntoXAI leverages semantic technology and machine learning (ML) to enhance prediction accuracy and generate user-comprehensible natural language explanations in the context of dengue disease classification. OntoXAI can be summarized into three key aspects. (1) Creates a knowledge base that incorporates domain-specific knowledge related to the disease. This allows for the integration of expert knowledge into the classification process. (2) OntoXAI presents a diagnostic classification system that utilizes patient symptoms as input to classify the disease accurately. By leveraging ML algorithms, it achieves high accuracy in disease classification. (3) OntoXAI introduces SWRL and ontology to integrate explainable AI techniques with Open AI API, enabling a better understanding of the classification process. By combining the power of machine learning algorithms with the ability to provide transparent, human-understandable explanations through Open AI API, this approach offers several advantages in enhancing prediction accuracy, achieving levels of up to 96%. Overall, OntoXAI represents a significant advancement in the field of explainable AI, addressing the challenges of transparency and trust in machine learning systems, particularly in critical domains like healthcare. [ABSTRACT FROM AUTHOR]

Ontology Learning aims to facilitate automatic or semi-automatic ontology development based on machine learning techniques in context of big data. Recent evolution of technology has introduced Generative Artificial Intelligence (AI) capable of creating new data, extracting insights from the existing data, and generating coherent texts from various inputs. This ability supports analysis of text data, providing insights and annotations that reduce human effort. This study explores the emerging field of Generative AI, specifically, Large Language Models for ontology learning. We conducted a survey of the current state of Generative AI research with focus on applicability and efficacy for ontology development tasks, and assessment of evaluation techniques. We discussed challenges related to explainability and interpretability of Generative AI and outlined directions for future research. [ABSTRACT FROM AUTHOR]

The fast growth of the web generates a significant amount of heterogeneous information such as images, text, audio, and video through various applications. These applications use different layouts to represent significant information. The layouts of table information are overloaded with anomalies that have given rise to intensive research into the semantification of web content and organizing tabular data for knowledge sharing and acquisition. Moreover, there are many anomalies present in tabular layouts that lead to the lack of semantic representation in tabular form and new challenges in data modeling. In this paper, we have discussed the various anomalies present in the tabular data that pertain to ontology learning and population tasks and provide the semantification of tabular data. To complete this task, (1) we provide the list of anomalies that pertain to semantification and provide the resolution to anomalies along with the semantification of tabular data, and (2) we have established the algorithm to interpret the table structure into a formal representation to analyze anomalies and provide the resolution. Furthermore, the proposed approach has been compared with existing approaches using ontology elements, the ability to resolve the anomalies, and the time complexity of the ontology population. [ABSTRACT FROM AUTHOR]

Classifying domain entities into their respective top-level ontology concepts is a complex problem that typically demands manual analysis and deep expertise in the domain of interest and ontology engineering. Using an efficient approach to classify domain entities enhances data integration, interoperability, and the semantic clarity of ontologies, which are crucial for structured knowledge representation and modeling. Based on this, our main motivation is to help an ontology engineer with an automated approach to classify domain entities into top-level ontology concepts using informal definitions of these domain entities during the ontology development process. In this context, we hypothesize that the informal definitions encapsulate semantic information crucial for associating domain entities with specific top-level ontology concepts. Our approach leverages state-of-the-art language models to explore our hypothesis across multiple languages and informal definitions from different knowledge resources. In order to evaluate our proposal, we extracted multi-label datasets from the alignment of the OntoWordNet ontology and the BabelNet semantic network, covering the entire structure of the Dolce-Lite-Plus top-level ontology from most generic to most specific concepts. These datasets contain several different textual representation approaches of domain entities, including terms, example sentences, and informal definitions. Our experiments conducted 3 study cases, investigating the effectiveness of our proposal across different textual representation approaches, languages, and knowledge resources. We demonstrate that the best results are achieved using a classification pipeline with a K-Nearest Neighbor (KNN) method to classify the embedding representation of informal definitions from the Mistral large language model. The findings underscore the potential of informal definitions in reflecting top-level ontology concepts and point towards developing automated tools that could significantly aid ontology engineers during the ontology development process. [ABSTRACT FROM AUTHOR]

Reference ontologies provide a shared vocabulary and knowledge resource for their domain. Manual construction and annotation enables them to maintain high quality, allowing them to be widely accepted across their community. However, the manual ontology development process does not scale for large domains. We present a new methodology for automatic ontology extension for domains in which the ontology classes have associated graph-structured annotations, and apply it to the ChEBI ontology, a prominent reference ontology for life sciences chemistry. We train Transformer-based deep learning models on the leaf node structures from the ChEBI ontology and the classes to which they belong. The models are then able to automatically classify previously unseen chemical structures, resulting in automated ontology extension. The proposed models achieved an overall F1 scores of 0.80 and above, improvements of at least 6 percentage points over our previous results on the same dataset. In addition, the models are interpretable: we illustrate that visualizing the model's attention weights can help to explain the results by providing insight into how the model made its decisions. We also analyse the performance for molecules that have not been part of the ontology and evaluate the logical correctness of the resulting extension. [ABSTRACT FROM AUTHOR]

Researchers are working towards automation of ontology construction to manage the ever-growing data on the web. Currently, there is a shift from the use of machine learning techniques towards exploration of deep learning models for ontology construction. Deep learning model are capable of extracting terms, entities, relations, and classifications, and perform axiom learning from the underutilized richness of web-based knowledge. There has been remarkable progress in automatic ontology creation using deep learning models since they can perform word embedding, long-term dependency acquisition, concept extraction from large corpora, and inference of abstracted relationships based on broad corpora. Despite their emerging importance, deep learning models remain underutilized in ontology construction, and there is no comprehensive review of their application in ontology learning. This paper presents a comprehensive review of existing deep-learning models for the construction of ontologies, the strength and the weaknesses presented by the deep learning models for ontology learning as well as promising directions to achieve a more robust deep learning models. The Deep Learning models reviewed include Recurrent Neural Networks (RNNs), Convolutional Neural Networks (CNNs), Long-Short Term Memory (LSTMs), and Gated Recurrent Unit (GRU) as well as their ensembles. While these traditional deep learning models have achieved great success, one of their limitations is that they struggle to understand the meaning and order of data in sequences. CNNs and RNN-based models such as LSTMs and GRUs can be computationally expensive due to their large number of parameters or complex gating mechanisms. Furthermore, RNN models suffer from vanishing gradients, making it difficult to learn long-term relationships in sequences. Additionally, RNN-based models process information sequentially, limiting their ability to take advantage of powerful parallel computing hardware, slowing down training and inference, especially for long sequences. Consequently, there has been a shift towards Generative Pre-Trained (GPT) models and Bidirectional Encoder Representations from Transformers (BERT) models. This paper also reviewed the GPT-3, GPT-4, and the BERT models for extracting terms, entities, relations, and classifications. While GPT models excel in contextual understanding and flexibility, they fall short when handling domain-specific terminology and disambiguating complex relationships. Fine-tuning and domain-specific training data could minimize these shortcomings, and further enhance the performance of GPT in term and relation extraction tasks. On the other hand, the BERT models excel in comprehending context-heavy texts, but struggles with higher-level abstraction and inference tasks due to a lack of explicit semantic knowledge, thus necessitating inference for unspecified relationships. The paper recommends further research on deep learning models for ontology alignment and merging. Also, the ensembling of deep learning models and the use of domain-specific knowledge for ontology learning require further research for ontology construction.

Ontologies play an important role in representing the semantics of data sources. Building an ontology as a representation of domain knowledge from available data sources is not a simple process, particularly when dealing with relational data, which remains prevalent in existing knowledge systems. In this study, we create an ontology from a relational database using objectrelational mapping (ORM) metadata as additional rules for mapping. Our method comprises two main phases: ontology schema construction using ORM metadata and the generation of ontology instances from the relational database. During the initial phase, we analyzed the ORM metadata to map it to an resource description framework schema (RDF(S))-OWL representation of the ontology. In the subsequent phase, we applied mapping rules to convert the relational database (RDB) data into ontological instances, which are then represented as RDF triples. Using ORM metadata, we enhance the accuracy of the resulting ontology, particularly in terms of extracting concepts and hierarchical relationships. This study contributes to the field of ontology learning by showcasing a novel approach that leverages ORM metadata to create ontologies from relational databases. [ABSTRACT FROM AUTHOR]

Ontology is a significant data model for identifying semantic information for Coronavirus disease (COVID-19) discovery. Large-scale biological and general datasets have recently been more readily available for study and to support the development of COVID-19 repositioning, but efficiently utilizing and standard metadata modeling of these datasets still remains challenging. Coronavirus disease is a deadly disease. Researchers have focused on the different aspects of COVID-19, such as studies about COVID-19, symptoms, treatment, prevention, resources, infected cases, patient information, global impact, and related research. Compilation and analysis of metadata by knowledge gathered in various field studies can help in efficient data sharing and better decision-making of data related to COVID-19. However, considering the heterogeneous nature of data sources, it is challenging and innovative to make the COVID-19 information human and machine understandable and to provide answers to user queries at all possible aspects of the COVID-19 pandemic. This paper propose a COVID-19 ontology named as CovidO metadata model that provides a common conceptual model to facilitate interoperability between metadata ontologies from various heterogeneous data sources. The CovidO has the following objectives: (1) The ontology serves as a reference schema for reporting COVID-19 data. (2) It also offer a customized schema of the existing ontologies to build a standard global data model. (3) It covers all the possible aspects/dimensions of COVID-19. (4) the proposed model aims to utilize findability, accessibility, interoperability, and reusability (FAIR) principles by providing common conceptualization metadata and thereby abstracting from heterogeneous structures of existing sources of (static) data. The experiments are conducted over CovidO using OOPS, OntoMetric, and RDF/SPARQL query to evaluate the efficiency of the proposed model. The result shows that the proposed model has a broad scope with fewer pitfalls than existing COVID-19 ontologies. [ABSTRACT FROM AUTHOR]

This paper presents a natural language-driven dialogue system designed to support healthcare professionals and students in the field of physical medicine and rehabilitation. The system seamlessly integrates concepts from intelligent information systems, data mining, ontologies, and human-computer interaction, employing at its core a rule-based dialogue mechanism. The system harnesses the power of ontology-based graph knowledge, underscoring its domain-specific efficacy. This article delves into the automated knowledge base formation, utilising Python scripts to translate EBSCO’s dataset of articles on physical medicine and rehabilitation into an OWL ontology. This methodology ensures adaptability to the ever-evolving landscape of medical insights. The system’s approach to natural language processing encompasses text preprocessing, semantic category discernment, and SPARQL query creation, providing 26 predefined categories. As an innovation in performance optimisation, the system integrates a strategy to cache precomputed responses using a PostgreSQL database, which aids in resource conservation and reduction in query execution latency. The system’s user engagement avenues are further detailed, showcasing a Telegram bot and an API, enhancing accessibility and user experience. In essence, this article illuminates an advanced, efficient dialogue system for physical medicine and rehabilitation, synthesising multiple computational paradigms, and standing as a beacon for healthcare practitioners and students alike. [ABSTRACT FROM AUTHOR]

A relational database (RDB) is a digital database that uses components (such as constraints, tables, keys, etc.) to manage data in a structured manner. Because of these components, RDBs are considered 'poor' from a semantic point of view, precisely because of the structure-oriented nature of the components used. One way to eliminate this limitation is to transform the RDB into an ontology. The purpose of this article is to review the different approaches existing in the literature to extract data from an RDB and convert it into ontology instances. Two approaches are used to integrate the mapping between RDBs and ontologies. The first allows ontologies to be extracted from an RDB, the second consists of a mapping of the relational database to an existing ontology. Our proposed review focuses on methods for creating a specific ontology from an RDB. The proposed review examines this field, classifying the methods that will be analyzed according to their inputs and outputs. Such classification may be useful for understanding the usability of methods. The aim is to critically review existing studies to help outline this research topic's progress and identify methods' gaps and functionalities. [ABSTRACT FROM AUTHOR]

This paper presents a natural language-driven dialogue system designed to support healthcare professionals and students in the field of physical medicine and rehabilitation. The system seamlessly integrates concepts from intelligent information systems, data mining, ontologies, and human-computer interaction, employing at its core a rule-based dialogue mechanism. The system harnesses the power of ontology-based graph knowledge, underscoring its domain-specific efficacy. This article delves into the automated knowledge base formation, utilising Python scripts to translate EBSCO’s dataset of articles on physical medicine and rehabilitation into an OWL ontology. This methodology ensures adaptability to the ever-evolving landscape of medical insights. The system’s approach to natural language processing encompasses text preprocessing, semantic category discernment, and SPARQL query creation, providing 26 predefined categories. As an innovation in performance optimisation, the system integrates a strategy to cache precomputed responses using a PostgreSQL database, which aids in resource conservation and reduction in query execution latency. The system’s user engagement avenues are further detailed, showcasing a Telegram bot and an API, enhancing accessibility and user experience. In essence, this article illuminates an advanced, efficient dialogue system for physical medicine and rehabilitation, synthesising multiple computational paradigms, and standing as a beacon for healthcare practitioners and students alike.

It is very difficult to predict stocks and commodity price index due to the presence of many and influential uncertainties. With the help of the accumulated information available in the current digital age and the power of high-performance computing machines, there is a lot of focus on designing algorithms that can learn stock market trends and successfully predict stock prices. Therefore, it will be very useful to create appropriate knowledge bases in order to increase the accuracy and efficiency of these systems and to facilitate the routine of using conventional knowledge in machine learning systems. The purpose of this research is to develop a Persian ontology for modeling the stock market and identifying factors affecting the stock market. The created ontology will lead to the enrichment and completion of the capacities of the existing knowledge bases in this field. For this purpose, in this research, a domain-specific ontology has been developed in the field of stock market and financial markets, which was prepared in Persian language by the authors of this research. After introducing this ontology, the details of the steps required to collect relevant data, semi-automated development and evaluation of this knowledge resource are described. The constructed ontology includes 565 concepts, 496 hierarchical relationships, 137 non-hierarchical relationships, and 937 samples that have been evaluated with various criteria and have a favorable status. It seems that this ontology in the current conditions and according to the evaluated volume and quality, is quite suitable to be used as a source of knowledge to improve the performance of machine learning systems for stock forecasting, and it can also be used to training stock market analysts and creating a knowledge base for brokerages, improving the process of retrieving semantic information and helping to determine the investment strategies of individuals in investment funds. [ABSTRACT FROM AUTHOR]

With the exponential increase of data on the web, the manual acquisition of ontology has become a time-consuming and tedious task. Thus, switching to ontology learning enables the ontologies' acquisition automation. In this paper, we deal with the phase of concepts' extraction. Our motivation is to capitalise on the confirmed advantages of deep learning (DL) models and word embedding techniques to automatically extract relevant concepts from large amounts of textual data. A four phases approach is proposed where different models and techniques are applied and a comparative study is achieved: the preprocessing phase, the classification phase, based on DL models, the terms filtering phase, where we experimented and compared three methods to extract the relevant terms, and the semantic enrichment phase experimenting and comparing word embedding techniques to semantically enrich the discovered concepts. The approach is implemented and evaluated on different medical datasets. The obtained results proved the suitability of the experimented models and techniques for the concepts' extraction. [ABSTRACT FROM AUTHOR]

A new technique and its software implementation are presented to create a deeply semantically structured ontology using plain natural language text as input, without regular structure or any previous tagging and markup. The new approach is primarily aimed at highly inflectional languages, and is implemented for Ukrainian. The automatically created ontologies (in OWL) could be easily converted to other graph databases formats, such as Neo4j, and were successfully evaluated as valid ontologies using Protégé, RDFlib and Neo4j environments. An integrated approach is proposed for the development of natural language dialogue systems driven by the ontologyrelated graph database using the Cypher language for the formal queries. The original phrases are subject to a special method of semantic analysis, which determines the type of formal query to the database. The essence of the analysis is that the text of the user’s phrase goes through a series of checks. Based on their results, a set of basic templates for the formal requests are determined, as well as additional constructions that are attached to the basic template. Some of the checks may also return the notion of substitution to certain specified positions of the formal query. Formal queries can return both contexts and lists of ontology concepts. In addition to concepts, queries can also return information about specific semantic predicates that connect them, which simplifies the synthesis of natural language responses. The synthesis of answers is based on special templates, the choice of which is directly related to the corresponding template of the formal query. [ABSTRACT FROM AUTHOR]

Since the beginning of the century, research on ontology learning has gained popularity. Automatically extracting and structuring knowledge relevant to a domain of interest from unstructured textual data is a major scientific challenge. After studying the main existing methods, such as Text2Onto, we propose a new approach with a modular ontology learning framework focusing on automatically extracting knowledge from raw text sources. We consider tasks from data pre-processing to axiom extraction. Whereas previous contributions considered ontology learning systems as tools to help the domain expert craft a reusable ontology, we developed the proposed framework with full automation in mind to build a minimum viable ontology targeted at an application. Ontology Learning Applied Framework (OLAF) has been generically designed to build specific ontologies whatever the application domain, use case and text data. We implement an initial version and test the framework on an ontology-based system, a search engine for technical products. [ABSTRACT FROM AUTHOR]

Knowledge base construction (KBC) using AI has been one of the key goals of this highly popular technology since its emergence, as it helps to comprehend everything, including relations, around us. The construction of knowledge bases can summarize a piece of text in a machine-processable and understandable way. This can prove to be valuable and assistive to knowledge engineers. In this paper, we present the application of natural language processing in the construction of knowledge bases. We demonstrate how a trained bidirectional long short-term memory or bi-LSTM neural network model can be used to construct knowledge bases in accordance with the exact ISO26262 definitions as defined in the GENIAL! Basic Ontology. We provide the system with an electronic text document from the microelectronics domain and the system attempts to create a knowledge base from the available information in textual format. This information is then expressed in the form of graphs when queried by the user. This method of information retrieval presents the user with a much more technical and comprehensive understanding of an expert piece of text. This is achieved by applying the process of named entity recognition (NER) for knowledge extraction. This paper provides a result report of the current status of our knowledge construction process and knowledge base content, as well as describes our challenges and experiences. [ABSTRACT FROM AUTHOR]

Purpose. During the development of railway ontologies, it is necessary to take into account both the data of information systems and regulatory support to check their consistency. To do this, data integration is performed. The purpose of the work is to formalize the methods for integrating heterogeneous sources of information and ontology formation. Methodology. Constructive-synthesizing modelling of ontology formation and its resources was developed. Findings. Ontology formation formalization has been performed, which allows expanding the possibilities of automating the integration and coordination of data using ontologies. In the future, it is planned to expand the structural system for the formation of ontologies based on textual sources of railway regulatory documentation and information systems. Originality. The authors laid the foundations of using constructive-synthesizing modelling in the railway transport ontological domain to form the structure and data of the railway train speed restriction warning tables (database and csv format), their transformation into a common tabular format, vocabulary, rules and ontology individuals, as well as ontology population. Ontology learning methods have been developed to integrate data from heterogeneous sources. Practical value. The developed methods make it possible to integrate heterogeneous data sources (the structure of the table of the railway train management rules, the form and application for issuing a warning), which are railway domain-specific. It allows forming an ontology from its data sources (database and csv formats) to schema and individuals. Integration and consistency of information system data and regulatory documentation is one of the aspects of increasing the level of train traffic safety.