1,721,047 research outputs found
Unlocking Historical Insights: Developing a Dataset from Historical Archives
No full textThe proliferation of data on the Web has resulted in an increased need for effective techniques to extract relevant and valuable knowledge from this data. The intersection of the fields of Information Extraction and Semantic Web has created new opportunities to improve ontology-based information extraction tools. However, the development and evaluation of such systems have been hampered by the scarcity of annotated documents, particularly in historical domains. This article discusses the current state of our work in creating a large RDF dataset that aims to support the development of ontology-based extraction tools. The dataset was created through manual annotation by domain experts as part of the arkivo project and contains approximately 300,000 triples, which are freely available. This dataset can be used as a benchmark to evaluate systems that automatically extract entities and annotate documents
Building the semantic layer of the Józef Piłsudski digital archive with an ontology-based approach
No full textUsing semantic web technologies is becoming an efficient way to overcome metadata storage and data integration problems in digital archives, thus enhancing the accuracy of the search process and leading to the retrieval of more relevant results. In this paper, the results of the implementation of the semantic layer of the Józef Piłsudski Institute of America digital archive are presented. In order to represent and integrate data about the archival collections housed by the institute, the authors developed arkivo, an ontology that accommodates the archival description of records but also provides a reference schema for publishing linked data. The authors describe the application of arkivo to the digitized archival collections of the institute, with emphasis on how these resources have been linked to external datasets in the linked data cloud. They also show the results of an experiment focused on the query answering task involving a state-of-the-art triple store system. The dataset related to the Piłsudski Institute archival collections has been made available for ontology benchmarking purposes
Verifying Neural Networks with Non-Linear SMT Solvers: a Short Status Report
No full textIn the last couple of decades, the popularity of neural networks has soared and they have been successfully utilized in many different domains across computer science. However, their application in safety and security-critical domains has been limited due to concerns regarding their reliability. Traditional methods for verifying neural networks (NNs) often uses linear Satisfiability Modulo Theory (SMT) solvers. These solvers work well for simple and shallow NN architectures but face limitations regarding their inability to handle non-linear activations, pooling layers, and complex activation functions, commonly used in modern deep neural networks.In this paper, we explore the potential of non-linear SMT solvers to verify intricate neural network architectures. By leveraging non-linear SMT solvers, a wider range of activation functions can be considered, leading to more accurate reasoning about the behavior of complex deep neural networks. The focus is on using recent advancements in SMT solver development to verify NNs with non-linear activation functions, particularly in the context of Computer Vision tasks. To test this idea, we conducted an experimental analysis to assess whether current nonlinear SMT solvers can efficiently handle NNs with transcendent activation functions
Constructing a Knowledge Graph for Italian Cinema Divas' Autobiographies
No full textAutobiographical writings are invaluable for research, offering relevant insights into personal experiences and cultural contexts. This is particularly true for Italian actresses, whose autobiographies, while rich with information, have been relatively underexplored in academic research. The Women Writing around the Camera (WOW) project addresses this gap by developing a semantic portal dedicated to these autobiographical texts. The WOW portal will reveal the dynamics between the actresses' writings, their private lives, their artistic careers, and the cultivation of the diva image. As a first step towards this goal, this paper presents the WOW knowledge graph (KG), which maps the personal and professional networks related to the divas' lives. The KG was built starting from the actresses' autobiographies, guided by a taxonomy of themes curated by domain experts. Although still under development and expansion, the KG provides a solid foundation for future enhancements
Detection of Component Degradation: A Study on Autoencoder-Based Approaches
No full textIn the realm of predictive maintenance, the incorporation of artificial intelligence (AI) methods has revolutionized the field by empowering businesses to actively monitor and preemptively address equipment malfunctions. Detecting anomalies plays a crucial role in predictive maintenance as it serves as an early indicator of potential faults or failures. This paper introduces initial findings from the use of autoencoders and their associated vector reconstruction error within the context of the IMOCO4.E project
Verifying Neural Networks with SMT: An Experimental Evaluation
No full textThe popularity of neural networks has grown significantly in various domains, however their use in safety-critical areas has been restricted due to reliability concerns. The AIDOaRt project, an H2020-ECSEL European initiative, aims to develop dependable neural networks for safety-critical contexts. This work investigates the application of Satisfiability Modulo Theory technologies to verify neural networks with non-linear activation functions in computer vision tasks
Verification of NNs in the IMOCO4.E Project: Preliminary Results
No full textIn recent years, there has been growing interest in machine learning and neural networks within research and industrial communities. While neural networks have shown impressive capabilities across various domains, their practical applications are still limited in safety-critical contexts due to a lack of formal guarantees regarding their reliability and behavior. This paper explores the latest advancements in Satisfiability Modulo Theory (SMT) technologies for verifying neural networks with piece-wise linear and transcendent activation functions. Through experimental analysis, we evaluate these technologies using neural networks trained on a real-world predictive maintenance dataset. This research contributes to the ongoing efforts to enhance the safety and reliability of neural networks through formal verification, enabling their deployment in safety-critical domains
Anomaly Recognition with Trustworthy Neural Networks: a Case Study in Elevator Control
No full textIn the realm of contemporary industrial control systems, the necessity for robust anomaly detection and classification is of critical importance. This paper presents an application of neural network technology in a real-world industrial scenario focused on elevator control. We employ two fully-connected neural networks to accomplish both anomaly detection and classification. The first neural network is dedicated to identifying types of anomalies, while the second predicts their magnitudes. Additionally, we integrate formal verification to certify the local robustness of these networks. Our findings not only showcase the practical efficacy of our methodology but also emphasise the crucial role of small neural networks in effectively addressing challenges within industrial settings
ARKIVO: An ontology for describing archival resources
No full textIn this paper we present ARKIVO, an ontology designed to accommodate the archival description of historical document collections. The aim of ARKIVO is to provide a reference schema for a rich representation of data elements in digital historical archives. This paper briefly reports design and implementation of ARKIVO, as well as its application on a real world case study, namely the Jozef Pilsudski Institute of America digitized collections. © Copyright 2018 for the individual papers by the papers' authors
- …
