Search

Recently, quantum computing has been proven as an ideal theory for the design of fuzzy inference engines, thanks to its capability to efficiently solve the rule explosion problem. In this scenario, a quantum fuzzy inference engine (QFIE) was proposed as a quantum algorithm able to generate an exponential computational advantage over conventional fuzzy inference engines. However, there are no practical demonstrations that the QFIE can be used to efficiently manage complex systems. This article bridges this gap by using, for the very first time, the QFIE to control critical systems such as those related to particle accelerator facilities at the European Organization for Nuclear Research (CERN). As demonstrated by a series of experiments performed at the T4 target station of the CERN Super Proton Synchrotron fixed-target physics beamline and at the Advanced Proton Driven Plasma Wakefield Acceleration Experiment, the QFIE is able to efficiently control such an environment, paving the way for the use of fuzzy-enabled quantum computers in real-world applications.

Evolutionary Algorithms (EAs) are becoming increasingly popular for training Variational Quantum Circuits (VQCs) due to their ability to conserve quantum resources. However, there is currently a lack of user-friendly tools for implementing this approach. To address this issue, this paper proposes EVOVAQ, a Python-based framework designed to simplify the use of EAs for training VQCs. EVOVAQ seamlessly integrates evolutionary computation with quantum libraries such as Qiskit, making it easy to use for both quantum computing and EAs communities. Furthermore, EVOVAQ’s scalability enables the development of customized solutions, promoting innovation in the quantum computing field.

Nowadays, machine learning is being used to address multiple problems in various research fields, with software engineering researchers being among the most active users of machine learning mechanisms. Recent advances revolve around the use of quantum machine learning, which promises to revolutionize program computation and boost software systems’ problem-solving capabilities. However, using quantum computing technologies is not trivial and requires interdisciplinary skills and expertise. For such a reason, we propose QuantuMoonLight, a community-based low-code platform that allows researchers and practitioners to configure and experiment with quantum machine learning pipelines, compare them with classic machine learning algorithms, and share lessons learned and experience reports. We showcase the architecture and main features of QuantuMoonLight, other than discussing its envisioned impact on research and practice.

Twitter is currently one of the most popular platforms for disseminating information about events happening around the world. Especially but not only for emergency events, it is crucial to know when and where the events are taking place. Unfortunately, identifying the geo-location of an event discussed in Twitter is a very challenging task mainly due to the brevity of the messages (i.e., tweets) and their subjective nature. In the literature, some efforts have been made to address this task, but they are characterized by substantial limitations such as the use of exclusively text analysis techniques, or the need for keywords or possible candidate locations. This paper proposes a new process for automatic event geo-localization which relies on both textual and spatial/temporal use of content posted on Twitter without using some prior knowledge about the event to be located. As shown by experimental results, our proposal achieves a good accuracy rate and outperforms two well-known baseline approaches related to the geo-location of events in Twitter.

Recognizing and collecting evidence at a crime scene are essential tasks for gathering information about perpetrators and/or the dynamics of a criminal event. Hence, the success of a crime investigation is strongly based on the ability of forensic investigators to perform these tasks. Recent studies observing and comparing the performance of experts and novices have highlighted the importance of experience and training for search and recovery strategies at crime scenes. Therefore, relevant training programs in evidence-recovery techniques should be attended by novices to improve their skills. However, the knowledge transfer between skills acquired in the classroom and their practical application in the field is a challenging task. In order to relieve this problem, this paper proposes a serious mixed-reality game, which is called TraceGame, aiming to support the training activities of novice forensic investigators by improving their skills related to the search and recovery of evidence at crime scenes. The purpose of the game is to identify the greatest number of useful traces present in a crime scene that is physically reconstructed at the training site as quickly as possible. As shown in an experimental session, TraceGame is a promising tool for supporting the training of novice forensic investigators.

Quantum computing is rapidly establishing itself as a new computing paradigm capable of obtaining advantages over its classical counterpart. However, a major limitation in the design of a quantum algorithm is related to the proper mapping of the corresponding circuit to a specific quantum processor so that the underlying physical constraints are satisfied. Moreover, current deterministic mapping algorithms suffer from high run times as the number of qubits to map increases. To bridge the gap in view of the next generation of quantum computers composed of thousands of qubits, this data paper proposes the first datasets that help address the quantum circuit mapping problem as a classification task. Each dataset is composed of random quantum circuits mapped onto a specific IBM quantum processor. In detail, each dataset instance contains some features related to the calibration data of the physical device and others related to the generated quantum circuit. Finally, the instance is labeled with a vector encoding the best mapping among those provided by deterministic mapping algorithms. Considering this, the proposed datasets allow the development of machine learning models capable of achieving mapping similar to those achieved with deterministic algorithms, but in a significantly shorter time.

In the age of the Internet of Things, online data have witnessed a significant growth in terms of volume and diversity, and research into information retrieval has become one of the important research themes in the Internet-oriented data science research. This paper introduces a novel domain knowledge centric methodology aimed at improving the accuracy of using machine learning methods for relation extraction from text data, which is critical to the accuracy and efficiency of information retrieval-based applications, including recommender systems and sentiment analysis. The proposed methodology makes a significant contribution to the processes of domain knowledge-based relation extraction including interrogating Linked Open Datasets to generate the relation classification training data, addressing the imbalanced classification in the training datasets, determining the probability threshold of the best learning algorithm, and establishing the optimum parameters for genetic algorithms, which were utilized to optimize the feature selection for the learning algorithms. The experimental evaluation of the proposed methodology reveals that the adopted machine-learning algorithms exhibit higher precision and recall in relation extraction in the reduced feature space optimized by our implementation. The considered machine learning includes support vector machine, perceptron algorithm uneven margin, and K-nearest neighbors. The outcome is verified by comparing against the random mutation hill-climbing optimization algorithm using Wilcoxon signed-rank statistical analysis.

The paper presents a collection of electroencephalography (EEG) data from a portable Steady State Visual Evoked Potentials (SSVEP)-based Brain Computer Interface (BCI). The collection of data was acquired by means of experiments based on repetitive visual stimuli with four different flickering frequencies. The main novelty of the proposed data set is related to the usage of a single-channel dry-sensor acquisition device. Different from conventional BCI helmets, this kind of device strongly improves the users’ comfort and, therefore, there is a strong interest in using it to pave the way towards the future generation of Internet of Things (IoT) applications. Consequently, the dataset proposed in this paper aims to act as a key tool to support the research activities in this emerging topic of human-computer interaction.

The need for widespread structural safety checks represents a stimulus for the research of advanced techniques for structural monitoring at the scale of single constructions or wide areas. In this work, a strategy to preliminarily identify and rank possible critical constructions in a built environment is presented, based on the joint exploitation of satellite radar remote sensing measurements and artificial intelligence (AI) techniques. The satellite measurements are represented by the displacement time series obtained through the Differential Synthetic Aperture Radar Interferometry (DInSAR) technique known as full resolution Small BAseline Subset (SBAS) approach, while the exploited AI technique is represented by the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) methodology. The DBSCAN technique is applied to the SBAS-DInSAR products relevant to the achieved Persistent Scatterers (PSs), to identify clusters of pixels corresponding to buildings within the investigated area. The analysis of the deformation evolution of each building cluster is performed in terms of velocity rates and statistics on the DInSAR measurements. Synthetic deformation maps of the areas are then retrieved to identify critical buildings. The proposed methodology is applied to three areas within the city of Rome (Italy), imaged by the COSMO-SkyMed SAR satellite constellation from ascending and descending orbits (in the time interval 2011–2019). Starting from the DInSAR measurements, the DBSCAN algorithm provides the automatic clustering of buildings within the three selected areas. Exploiting the derived deformation maps of each study area, a preliminary identification and ranking of critical buildings is achieved, thus confirming the validity of the proposed approach.

Fuzzy logic systems are useful for solving problems in many application fields. However, these systems are usually stored in specific formats and researchers need to rewrite them to use in new problems. Recently, the IEEE Computational Intelligence Society has sponsored the publication of the IEEE Standard 1855-2016 to provide a unified and well-defined representation of fuzzy systems for problems of classification, regression, and control. The main aim of this standard is to facilitate the exchange of fuzzy systems across different programming systems in order to avoid the need to rewrite available pieces of code or to develop new software tools to replicate functionalities that are already provided by other software. In order to make the standard operative and useful for the research community, this paper presents JFML, an open source Java library that offers a complete implementation of the new IEEE standard and capability to import/export fuzzy systems in accordance with other standards and software. Moreover, the new library has associated a Website with complementary material, documentation, and examples in order to facilitate its use. In this paper, we present three case studies that illustrate the potential of JFML and the advantages of exchanging fuzzy systems among available software.

The Internet of Things (IoT) paradigm is establishing itself as a technology to improve data acquisition and information management in the construction field. It is consolidating as an emerging technology in all phases of the life cycle of projects and specifically in the execution phase of a construction project. One of the fundamental tasks in this phase is related to Health and Safety Management since the accident rate in this sector is very high compared to other phases or even sectors. For example, one of the most critical risks is falling objects due to the peculiarities of the construction process. Therefore, the integration of both technology and safety expert knowledge in this task is a key issue including ubiquitous computing, real-time decision capacity and expert knowledge management from risks with imprecise data. Starting from this vision, the goal of this paper is to introduce an IoT infrastructure integrated with JFML, an open-source library for Fuzzy Logic Systems according to the IEEE Std 1855-2016, to support imprecise experts’ decision making in facing the risk of falling objects. The system advises the worker of the risk level of accidents in real-time employing a smart wristband. The proposed IoT infrastructure has been tested in three different scenarios involving habitual working situations and characterized by different levels of falling objects risk. As assessed by an expert panel, the proposed system shows suitable results.

The prediction of cancer staging in prostate cancer is a process for estimating the likelihood that the cancer has spread before treatment is given to the patient. Although important for determining the most suitable treatment and optimal management strategy for patients, staging continues to present significant challenges to clinicians. Clinical test results such as the pre-treatment Prostate-Specific Antigen (PSA) level, the biopsy most common tumor pattern (Primary Gleason pattern) and the second most common tumor pattern (Secondary Gleason pattern) in tissue biopsies, and the clinical T stage can be used by clinicians to predict the pathological stage of cancer. However, not every patient will return abnormal results in all tests. This significantly influences the capacity to effectively predict the stage of prostate cancer. Herein we have developed a neuro-fuzzy computational intelligence model for classifying and predicting the likelihood of a patient having Organ-Confined Disease (OCD) or Extra-Prostatic Disease (ED) using a prostate cancer patient dataset obtained from The Cancer Genome Atlas (TCGA) Research Network. The system input consisted of the following variables: Primary and Secondary Gleason biopsy patterns, PSA levels, age at diagnosis, and clinical T stage. The performance of the neuro-fuzzy system was compared to other computational intelligence based approaches, namely the Artificial Neural Network, Fuzzy C-Means, Support Vector Machine, the Naive Bayes classifiers, and also the AJCC pTNM Staging Nomogram which is commonly used by clinicians. A comparison of the optimal Receiver Operating Characteristic (ROC) points that were identified using these approaches, revealed that the neuro-fuzzy system, at its optimal point, returns the largest Area Under the ROC Curve (AUC), with a low number of false positives (FPR = 0.274, TPR = 0.789, AUC = 0.812). The proposed approach is also an improvement over the AJCC pTNM Staging Nomogram (FPR = 0.032, TPR = 0.197, AUC = 0.582).