From implicit semantics towards ontologies — practical considerations from the INTER-IoT perspective

In this paper, we tackle the problem of deploying face recognition (FR) solutions in heterogeneous Internet of Things (IoT) platforms. The main challenges are the optimal deployment of deep neural networks (DNNs) in the high variety of IoT devices (e.g., robots, tablets, smartphones, etc.), the secure management of biometric data while respecting the users’ privacy, and the design of appropriate user interaction with facial verification mechanisms for all kinds of users. We analyze different approaches to solving all these challenges and propose a knowledge-driven methodology for the automated deployment of DNN-based FR solutions in IoT devices, with the secure management of biometric data, and real-time feedback for improved interaction. We provide some practical examples and experimental results with state-of-the-art DNNs for FR in Intel’s and NVIDIA’s hardware platforms as IoT devices.

NoSQL and relational databases have been integrated in response to the increasing demand for scalable and efficient data management in hybrid cloud environments. The differences in data structures and query processing methods between these databases present both challenges and opportunities when designing an optimized hybrid system. This study explores the integration of NoSQL and relational databases to maximize data storage, retrieval, and processing efficiency across multiple cloud systems. NoSQL databases excel in handling unstructured and semi-structured data, offering flexibility and scalability, whereas relational databases provide robust consistency and structured query capabilities. Despite the wide availability of relational databases, certain applications require the dynamic adaptability of NoSQL systems, making integration a viable solution. The research evaluates key performance parameters, including query execution speed, scalability, data consistency, and resource utilization. Cloudsim is used for simulation, allowing for an in-depth comparison between standalone and hybrid database models. Experimental results indicate that the proposed hybrid model improves query performance by 30%, reduces latency by 20%, and enhances scalability by 40% compared to using relational or NoSQL databases alone. The novelty of this approach lies in its ability to overcome the limitations of each database type by their strengths in a dynamic cloud environment. The results highlight the effectiveness of hybrid database integration in optimizing cloud-based data management while ensuring seamless operation, adaptability, and resource efficiency.

In this paper, we present ML-Quadrat, an open-source research prototype that is based on the Eclipse Modeling Framework (EMF) and the state of the art in the literature of Model-Driven Software Engineering (MDSE) for smart Cyber-Physical Systems (CPS) and the Internet of Things (IoT). Its envisioned users are mostly software developers who might not have deep knowledge and skills in the heterogeneous IoT platforms and the diverse Artificial Intelligence (AI) technologies, specifically regarding Machine Learning (ML). ML-Quadrat is released under the terms of the Apache 2.0 license on Github <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> . Additionally, we demonstrate an early tool prototype of DriotData, a web-based Low-Code platform targeting citizen data scientists and citizen/end-user software developers. DriotData exploits and adopts ML-Quadrat in the industry by offering an ex-tended version of it as a subscription-based service to companies, mainly Small- and Medium-Sized Enterprises (SME). The current preliminary version of DriotData has three web-based model editors: text-based, tree-/form-based and diagram-based. The latter is designed for domain experts in the problem or use case domains (namely the IoT vertical domains) who might not have knowledge and skills in the field of IT. Finally, a short video demonstrating the tools is available on YouTube: https://youtu.be/VAuz25w0a5k.

Modern large-scale information systems often use multiple database management systems, not all of which are necessarily relational. In recent years, NoSQL databases have gained acceptance in certain domains while relational databases remain de facto standard in many others. Many "legacy" information systems also use relational databases. Unlike relational database systems, NoSQL databases do not have a common data model or query language, making it difficult for users to access data in a uniform manner when using a combination of relational and NoSQL databases or simply several different NoSQL database systems. Therefore, the need for uniform data access from such a variety of data sources becomes one of the central problems for data integration. In this paper we provide an overview of the main problems, methods, and solutions for data integration between relational and NoSQL databases, as well as between different NoSQL databases. We focus mainly on the problems of structural, syntactic, and semantic heterogeneity and on proposed solutions for uniform data access, emphasizing some of the more recent proposals.

This study intends to help data mining developers to get better performance when obtaining the most frequent N-grams in Text Mining projects. The process of building new variables is one of the oldest and still challenging problems in Data Mining projects. The most frequent N-grams are commonly used as input variables in Text Mining projects. The N-grams represent the occurrence of N items in sequence in a given text. The items can be letters or words. This paper presents a performance comparison between the two main approaches of data storage, relational and NoSQL databases in the task of obtaining the most frequent N-grams. Validation of the study was executed using a database from a known benchmark from an international competition organized by PAN@CLEF 2013. The one-tailed paired t-test showed that NoSQL approach is statistically superior to the relational approach with a confidence level of 95%.

The virtual representation and integration of the internet with the physical objects, devices or things have been growing exponentially in recent years. This has motivated the community to design and develop new Internet of Things (IoT) platforms to cater, capture, access, store, share, and communicate data for information retrieval and intelligent applications. However, the associated dynamism, resource-constrain, cost and the nature of the IoT warrants special design obligations for its effectiveness in the days ahead, hence pose a challenge to the community. The understanding of web data from machines according to the subject of terminology in different fields is a complex task. It opens up new challenges to researchers as such an effort mandates the provision of semantically structured, appropriate information sources in this information age. The advent of numerous smart devices, operators, and IoT service providers subject to time-consuming and complex operations, inadequate research and innovations give rise to design complexity. For efficient functioning and effective implementation of the domain requires the inclusion of semantics and the desired interoperability among these factors. This motivates the authors to review and emphasizes a few of the emerging trends of the semantic technology impacting the IoT. Particularly, the work focuses on different aspects as information modeling, ontology design, machine learning, network tools, security policy and processing of semantic data—and discuss the issues and challenges in the current scenario.

The project entitled as Big Data, Internet of Things, and Mobile Devices, in Portuguese Banco de Dados, Internet das Coisas e Dispositivos Moveis (BDIC-DM) was implemented at the Brazilian Aeronautics Institute of Technology (ITA) on the 1st Semester of 2015. It involved 60 graduate students within just 17 academic weeks. As a starting point for some features of real time Online Transactional Processing (OLTP) system, the Relational Database Management System (RDBMS) MySQL was used along with the NoSQL Cassandra to store transaction data generated from web portal and mobile applications. Considering batch data analysis, the Apache Hadoop Ecosystem was used for Online Analytical Processing (OLAP). The infrastructure based on the Apache Sqoop tool has allowed exporting data from the relational database MySQL to the Hadoop File System (HDFS), while Python scripts were used to export transaction data from the NoSQL database to the HDFS. The main objective of the BDIC-DM project was to implement an e-Commerce prototype system to manage credit card transactions, involving large volumes of data, by using different technologies. The used tools involved generation, storage, and consumption of Big Data. This paper describes the process of integrating NoSQL and relational database with Hadoop Cluster, during an academic project using the Scrum Agile Method. At the end, processing time significantly decreased, by using appropriate tools and available data. For future work, it is suggested the investigation of other tools and datasets.

The main problem has been managing databases because they are continuously expanding quickly and getting more complicated in terms of volume, variety, and velocity. Currently, Relational Database Management Systems (RDMS), also known as SQL, or traditional search engines are primarily used to manage and utilize data collection. Due to their highly semantic properties and applications, relational databases have emerged as powerful and superior during the previous two decades. The handling of a high volume of data and the variability in data type and structure has become a laborious task since Big Data entered the IT industry. Relational databases are inadequate to handle such big data due to their rigid data limitations, structure, relations, and other factors. As a result, data aggregation is rendered impossible. NoSQL databases offer a practical and understandable foundation for combining massive amounts of data, structures, and interactions. Data modeling and migration are now needed in order to define the issue. There isn't yet a tool for switching from relational to no-SQL databases. For this migration, relational (SQL) database queries must be converted to NoSQL database queries. Since relational databases (RDBMS) have found it difficult to keep up with modernization, NoSQL has emerged as the most practical database. This essay aims to highlight and assess the applications of NoSQL, the notion of NoSQL data modeling, and the NoSQL database migration procedure. NoSQL, NoSQL Roadmap, Data Modeling, ETL Approach for Data Migration, Data Loading, and Data Extraction.

Recently, two terms, namely Big Data and Internet of Things (IoT) have gained popularity individually. However, their interconnections are not fully explored and understood. It is expected that the fusion of Big Data and IoT would create many complex systems for smart cities. While the data from IoT lies in Big Data, the scale of operations are completely different in terms of providing the required real-time analytics for such smart systems. Even though NoSQL databases and other next generation solutions could be deployed to achieve real-time responses, the major security challenges need to be understood as mission critical and sensitive data intertwines Big Data and IoT. In this paper, we identify the security challenges shared by the closely-knit Big Data and IoT in three main risk areas: i) NoSQL security vulnerabilities, ii) mobile IoT (M-IoT) security and privacy constraints and iii) encryption key security threats. We perform a comparative study of security vulnerabilities of NoSQL databases and identify the security and privacy constraints of M-IoT networks. Encryption key attacks for resource constrained IoT devices are also illustrated mathematically. Overall, this paper explores new research directions in these prime areas of security and privacy that would result in solution opportunities for a meaningful fusion of Big Data and IoT for a smart environment.

NoSQL databases boldly step into the area reserved so far for relational databases. NoSQL databases are more flexible than relational ones at cost of some disadvantages related especially with security and integrity constraints. The most convenient way to access a relational database in modern object applications are object-relational mapping tools. It turns out that some of these tools offer mappings to broad spectrum of NoSQL and object databases and text files. The aim of this article is to present this kind of tools and answer the question if they are really universal i.e. if they allow to persist, in any form, the same object entity as defined at the level of an object application.

Advances on networked ehealth information access and sharing: Status, challenges and prospects

The advancements concerning the development of ICT systems including Internet of Things (IoT), big data, cloud computing and NoSQL databases provide new opportunities and challenges for industrial asset management. The use of NoSQL databases has emerged due to the limitations of the relational databases, in particular, the inability to scale-up horizontally and to manage the data that is constantly generated by industry. The current work highlights the key aspects of both relational and NoSQL databases. The paper provides a review of the database technologies mentioned above. In this context, in order to demonstrate the effectiveness and adequacy of NoSQL databases, a real industrial case study is presented. The authors also discuss the different database technologies and their suitability in the domain of interest.

A data lake refers to an extremely large data resource or repository. Data lakes store large amounts of data and use advanced analytics to pair data from multiple sources with different types of structured, semi-structured, and unstructured information. NoSQL databases such as Mongodb, Redis, Neo4j, and Cassandra are nontabular and they store data differently rather than use relational tables. NoSQL databases come in many forms, mostly documents, key values, wide columns, and graphs based on their data model. NoSQL gives less complicated scalability and higher overall performance as compared with traditional relational databases. NoSQL databases can store different types of data, but they cannot fully support Automation, Consistency, Isolation, and endurance (ACID) features, i.e., trigger functions in multi-transaction management, because a NoSQL database uses a non-relational database system. Thus, an interpreter is necessary for SQL-to-NoSQL queries. We used ANTLR (ANother Tool for Language Recognition), which has five main stages: The input SQL query, the tokenizer, the parser, the parser tree, and lastly the generation of the query results in NoSQL. The tool gave users to write a flexible multi-pass language parser that is expected to solve problems in querying complex ACID functions and other problems in complex queries in NoSQL databases. In the measurement, analysis, and evaluation of the translation results through the comparison of each NoSQL criterion against a Relational Database Management System (MySQL), the scores obtained were as follows. The performance criterion achieved the highest score (98.40%) by the MongoDB database, followed by scalability (97.40%) and accuracy (97.00%). The criterion with the lowest score was complexity (91.65%).

The article focuses on comparing relational and non-relational databases using applications written in C#. The aim of the study is to understand in which cases relational databases are preferred and when it is worth considering the use of non-relational databases. The research examines the speed of data retrieval, updating, and deletion, in the context of five different databases, including relational ones like PostgreSQL, MySql, Oracle, and non-relational ones such as Neo4j and MongoDB. The data consists of 1,578,098 records. In the case of relational databases, a unified database model was applied, while in NoSQL databases, the data model was appropriately adjusted to the specific type of non-relational database. Differences in the execution time of database queries are analyzed based on indexing strategies and query complexity. Special attention is given to query performance efficiency in the context of using C# and libraries that facilitate the connection between the application and databases. The conducted research indicates that the PostgreSQL database achieves the lowest average query response times.

NoSQL database solutions are becoming more and more prevalent in a world currently dominated by SQL relational databases. NoSQL databases were designed to provide database solutions for large volumes of data that is not structured. However, the advantages (or disadvantages) of using a NoSQL database for data that is structured, and not necessarily Big, is not clear. There are not many studies that compare the performance of processing a modest amount of structured data in a NoSQL database with a traditional relational database. In this paper, we compare one of the NoSQL solutions, MongoDB, to the standard SQL relational database, SQL Server. We compare the performance, in terms of runtime, of these two databases for a modest-sized structured database. Results show that MongoDB performs equally as well or better than the relational database, except when aggregate functions are utilized.