An ontology alignment framework for data-driven microservices

The paper reports on a case study regarding the integration of heterogeneous data sources, coming from different institutions, needed to support several studies related to the health status of the population of L’Aquila (Italy) after the earthquake of April 6th 2009. In detail, the paper initially describes all the health studies, then the data sources required to support them, finally proposes a simplified federated architecture and a straightforward technological solution to implement it.

A method and software framework for enriching private biomedical sources with data from public online repositories

In recent years, research on digital twins in the construction sector has grown significantly, particularly since 2018. However, most studies focus on isolated lifecycle phases and give limited attention to developing practices that address all stages of the building lifecycle. Current applications of digital twins in construction are largely concentrated on design and engineering, where Building Information Modeling (BIM) serves as the primary input. This dominance stems from BIM's role as an existing digital representation of built assets, which can be extended into a digital twin. One of the major barriers to wider implementation lies in data integration and interoperability—specifically, the challenge of consolidating information from diverse sources, formats, and lifecycle stages. Future research should prioritize the standardization of data formats, protocols, and application programming interfaces (APIs) to facilitate seamless exchange across digital twin systems. Additionally, the development of advanced data governance frameworks is essential to ensure accuracy, completeness, and security of information. Greater exploration into semantic data modeling and ontologies will also support the integration of heterogeneous data sources. Beyond technical issues, it is equally important to investigate the costs, benefits, and potential drawbacks of adopting digital twins across the entire building lifecycle. Expanding the scope of applications beyond design and engineering to include construction, operation and maintenance, as well as demolition and recovery phases, will be crucial. This paper therefore outlines the current state of digital twin research in construction, identifies critical challenges, and highlights key directions for future development.

Ontologies enable explicit expression of collective concepts and support Machine-to-Machine (M2M) interactions at the semantic level. Ontologies expressed in a standard language, such as the Web Ontology Language (OWL) and exposed on a network offer the potential for unprecedented interoperability solutions since they are semantically rich, computer interpretable and inherently extensible. In this chapter, we describe how we applied ontologies in OWL for rapid enterprise integration of heterogeneous data sources to track objects in a battlespace. We found that once a robust foundational domain ontology is established, it is easy and quick to integrate new data sources and therefore rapidly provide new system capabilities. In particular, we demonstrate how moving tracks can be quickly integrated with intelligence and space events to provide enhanced situational awareness using ontologies. This chapter also describes the overall SEER and SWORIER systems we developed, the latter of which translated OWL ontologies (and RDF instances) and Semantic Web Rule Language (SWRL) rules into Prolog, applied knowledge compilation techniques, and then at runtime, utilized a combined OWL/logic programming reasoned for efficient automated reasoning. We also briefly describe a more recent extension to the prototype and to the ontologies that we made to address more rigorous geospatial rules for unmanned autonomous vehicle (UAV) avoidance. Finally, we consider some issues raised by our work and future lines of research to address these.

BackgroundProtein-protein interaction (PPI) prediction remains a central task in systems biology to achieve a better and holistic understanding of cellular and intracellular processes. Recently, an increasing number of computational methods have shifted from pair-wise prediction to network level prediction. Many of the existing network level methods predict PPIs under the assumption that the training network should be connected. However, this assumption greatly affects the prediction power and limits the application area because the current golden standard PPI networks are usually very sparse and disconnected. Therefore, how to effectively predict PPIs based on a training network that is sparse and disconnected remains a challenge.ResultsIn this work, we developed a novel PPI prediction method based on deep learning neural network and regularized Laplacian kernel. We use a neural network with an autoencoder-like architecture to implicitly simulate the evolutionary processes of a PPI network. Neurons of the output layer correspond to proteins and are labeled with values (1 for interaction and 0 for otherwise) from the adjacency matrix of a sparse disconnected training PPI network. Unlike autoencoder, neurons at the input layer are given all zero input, reflecting an assumption of no a priori knowledge about PPIs, and hidden layers of smaller sizes mimic ancient interactome at different times during evolution. After the training step, an evolved PPI network whose rows are outputs of the neural network can be obtained. We then predict PPIs by applying the regularized Laplacian kernel to the transition matrix that is built upon the evolved PPI network. The results from cross-validation experiments show that the PPI prediction accuracies for yeast data and human data measured as AUC are increased by up to 8.4 and 14.9% respectively, as compared to the baseline. Moreover, the evolved PPI network can also help us leverage complementary information from the disconnected training network and multiple heterogeneous data sources. Tested by the yeast data with six heterogeneous feature kernels, the results show our method can further improve the prediction performance by up to 2%, which is very close to an upper bound that is obtained by an Approximate Bayesian Computation based sampling method.ConclusionsThe proposed evolution deep neural network, coupled with regularized Laplacian kernel, is an effective tool in completing sparse and disconnected PPI networks and in facilitating integration of heterogeneous data sources.

Ontology matching is an important process for integration of heterogeneous data sources. A large number of different matchers for comparing ontologies exist. They can be classified into element-level and structure-level matchers. The element-level matchers compare entities ignoring their relations with other entities, while the structure-level matchers consider these relations. The TF/IDF (term frequency / inverse document frequency) measure is useful for specifying key terms weights in documents. In our matching system we use the TF/IDF measure for comparing documents that store data about ontology entities. However, the TF/IDF does not take synonyms into account, and it may occur that the terms that describe two entities the best are synonyms. In this paper we propose a matcher that combines the TF/IDF measure with synonym recognition when determining key term weights, in order to improve the results of ontology matching. Evaluation of the matcher is performed on case study examples.

Data integration of heterogeneous sources is still an active research problem. This topic is especially crucial on enterprises that deal with new and different kinds of data as new services are provided. However, even though research efforts have been applied to the information integration, there is a lack of practical approaches ready to be implemented for real use cases. In this paper, we present a web-based framework and system built upon ontologies and other semantic techniques for the integration of heterogeneous data sources. The example scenario used to demonstrate the utility of the framework consists of a use case with the actual data of a telecom company.

Ontology matching is the process of finding semantic relationships between entities of different ontologies. It plays an important role in the integration of heterogeneous data sources. To find semantic correspondences between entities, various matchers can be used. When multiple matchers are integrated into a matching system, the issue of combining these matchers together and determining the importance of every basic matcher arises. In a parallel composition of basic matchers, several matchers are used independently and then their results are aggregated. In this paper, we propose a method for automatically determining weights of basic matchers in their parallel composition and the architecture of the matching system. An evaluation of the method is performed on case study examples concerning the matching of four ontology pairs and a comparison with other aggregation methods is given. Additionally, the architecture of the matching system based on our automatic weighted aggregation method is presented.

Businesses find it hard to combine data from different systems, formats, and sources. Some examples of these sources are structured data in relational databases, semi-structured data like JSON or XML, and unstructured data like text or multimedia files. Businesses need to be able to handle and mix all of these different types of data in order to get the most out of them. You can now utilize an architecture to put together data from languages that are distinct. This framework is meant to help you find a way to handle a lot of various types of data and sources without slowing down or losing consistency. Cloud storage, APIs, and machine learning are just a few of the cutting-edge technologies that the framework leverages to make sure that data systems can work together without any hassles. It allows businesses to combine their data while ensuring that all of their systems are accurate and high-quality. The platform also solves the problem of scalability, which means that companies can simply handle more and more data without any issues or delays. This plan helps businesses maintain better track of their data, which makes it easier to put together and less likely to make mistakes. Companies should always be able to see their data in the same way, no matter where it comes from or what format it is in. This makes it easier to choose. The answer also helps with data governance since it allows people ways to keep track of data history, set security standards, and make sure they follow the rules. When you have to interact with diverse kinds of data, the polyglot data integration framework is an excellent method to deal with problems that come up. In a world where data is king, it helps businesses get the most out of their data, keep things running smoothly, and remain ahead of the competition

In an era where data is the cornerstone of decision-making, organizations face the challenge of integrating diverse data sources from heterogeneous environments. Traditional Extract, Transform, and Load (ETL) processes focus primarily on syntactic transformations, ensuring compatibility between disparate datasets. However, these approaches fall short in addressing semantic inconsistencies that arise from differences in meaning, context, and relationships among data [6][7]. This paper proposes a semantic-aware ETL framework that integrates ontology-based reasoning and dynamic rule generation to bridge this gap. The framework enhances data coherence and alignment by leveraging cutting-edge semantic technologies, including Apache Jena for ontology management, SPARQL for semantic querying, and Protégé for ontology design. Detailed analysis, design, and evaluation of the framework showcase its ability to revolutionize ETL processes in data warehousing, enabling robust integration of heterogeneous data sources.

Large-scale energy infrastructure projects such as oil and gas developments, power generation facilities, and transmission networks are characterized by high hazard intensity, complex socio-technical interactions, and multi-contractor delivery environments. These conditions create persistent challenges for effective occupational safety risk control, particularly when traditional safety management approaches remain reactive, compliance-driven, and reliant on lagging indicators. In response, this paper develops a conceptual model for data-driven occupational safety risk control tailored to the operational and governance realities of large-scale energy projects. The proposed model conceptualizes safety risk as a dynamic and measurable system property that can be proactively managed through the systematic integration of heterogeneous data sources. These include operational process data, workforce and competency records, incident and near-miss reports, environmental conditions, and real-time monitoring technologies. The model is structured around four interrelated layers: (1) risk sensing and data acquisition, (2) analytics and risk interpretation, (3) decision support and governance integration, and (4) risk control and intervention mechanisms. Emphasis is placed on the use of leading indicators, predictive analytics, and feedback loops to enable early identification of emerging hazards and weak signals across project phases. By explicitly linking data-driven insights to decision authority, escalation pathways, and safety governance structures, the model addresses a critical gap between analytical capability and practical risk control. It also highlights the importance of organizational learning, continuous model recalibration, and cross-contractor data integration in complex project ecosystems. The conceptual contribution of this work lies in framing occupational safety management not merely as an operational function, but as an adaptive, intelligence-enabled control system embedded within project governance. The model provides a foundation for future empirical research and offers practical guidance for project owners, EPC contractors, and regulators seeking to enhance safety performance, improve risk predictability, and strengthen resilience in large-scale energy infrastructure delivery.

Integration of heterogeneous data sources is a crucial step in big data analytics, although it creates ambiguity issues during mapping between the sources due to the variation in the query terms, data structure and granularity conflicts. However, there are limited researches on effective big data integration to address the ambiguity issue for big data analytics. This paper introduces a self-adaptive model for big data integration by exploiting the data structure during querying in order to mitigate and resolve ambiguities. An assessment of a preliminary work on the Geography and Quran dataset is reported to illustrate the feasibility of the proposed model that motivates future work such as solving complex query.

Accurate real-time prediction of surface ozone at high spatial and temporal resolution is critical for air-quality management, exposure assessment, and public health protection. However, developing an operational hourly ozone prediction system at the national scale remains challenging due to the need for continuous data acquisition, integration of heterogeneous data sources, and computationally efficient modeling frameworks. This study presents a real-time, data-driven framework for high-resolution hourly ozone prediction across South Korea by constructing an automated, nationwide atmospheric database that integrates satellite, meteorological, and ground-based observations.We establish an operational data pipeline that collects and processes near-real-time meteorological observations from the Korea Meteorological Administration’s Operational Data Assimilation and Model (ODAM), atmospheric information derived from the GK2A geostationary satellite, and surface ozone measurements from the AIRKOREA monitoring network. All data streams are ingested on an hourly basis and systematically harmonized through temporal synchronization and spatial alignment to generate high-resolution predictors covering the entire Korean Peninsula. This integrated database enables consistent representation of rapidly evolving meteorological conditions, atmospheric composition, and surface-level air quality.Using the constructed real-time database, we develop a data-driven prediction model to estimate hourly surface ozone concentrations, with AIRKOREA observations used as the target variable. The modeling framework is designed with operational feasibility in mind, supporting continuous updates, automated preprocessing, and near-real-time inference without reliance on computationally expensive chemical transport models. The resulting system provides high-resolution ozone predictions that capture fine-scale spatiotemporal variability at the national level.Model evaluation demonstrates that integrating geostationary satellite data with real-time meteorological and surface observations substantially enhances the prediction of hourly ozone variability compared to single-source or static-input approaches. The proposed framework highlights the advantages of real-time, high-resolution, and nationwide data integration for operational ozone forecasting in South Korea. Beyond ozone, this scalable and extensible system provides a foundation for real-time prediction of additional atmospheric pollutants and supports the development of next-generation data-driven air-quality forecasting services.

The creation of unified, open, secure, reliable, and agile data spaces is essential for collecting, storing, and sharing data in a standardized and accessible manner, promoting data reuse and addressing current interoperability limitations. In this context, this research presents a proof of concept for a unified agronomic data space based on the structured integration of heterogeneous open data sources. The central hypothesis is that the automated acquisition, preprocessing, and harmonization of publicly available agronomic data can significantly improve accessibility, usability, and interoperability for agricultural decision support applications. To this end, a comprehensive analysis of relevant open data sources was conducted, followed by the design and implementation of configurable algorithms for automated data downloading, cleaning, validation, and integration. The proposed approach explicitly addresses key challenges such as heterogeneous data formats, inconsistent spatial and temporal resolutions, missing values, and outlier detection. As a result, a unified access point was developed, providing reliable agronomic information, including (i) preprocessed climatological time series, (ii) crop and phytosanitary data, (iii) high-resolution aerial orthophotography, (iv) remote-sensing imagery, (v) pest-related information, and (vi) time series of major vegetation indices. The proof of concept was implemented for olive groves in the Andalusian region of Spain; however, the methodology is fully transferable to other crops, regions, and institutional contexts where comparable open data sources are available. The results demonstrate the potential of shared agronomic data spaces to enhance data reuse, support scalable analytics, and facilitate interoperable, data-driven agricultural management beyond the specific regional case study.

The multi-national MIDAS (Meaningful Integration of Data Analytics and Services) project is developing a big data platform to facilitate the utilisation of a wide range of health and social care data. The platform will enable the integration of heterogeneous data sources, providing privacy-preserving analytics, forecasting tools and bespoke visualisations of actionable epidemiological data. An evaluation framework starting with a logic model and using the principles of realist evaluation developed working with users, and software developers. The tools used are a series of parallel case studies to address the requirements of stakeholder groups at critical time points during the project to ensure IT systems development is in line with user’s requirements. The process includes longitudinal interviews with stakeholders, regular feedback to users and developers, and measurement of stakeholder’s attitudes to the project using Q-methodology.