Abstract The frequency of wildfires increases yearly and poses a constant threat to the environment and human beings. Different factors, for example surrounding infrastructure to an area (e.g., campfire sites or power lines) contribute to the occurrence of wildfires. In this paper, we propose using a Spatio-Temporal Knowledge Graph (STKG) based on OpenStreetMap (OSM) data for modeling such infrastructure. Based on that knowledge graph, we use the RDF2vec approach to create embeddings for predicting wildfires, and we align different vector spaces generated at each temporal step by partial rotation. In an experimental study, we determine the effect of the surrounding infrastructure by comparing different data composition strategies, which involve a prediction based on tabular data, a combination of tabular data and embeddings, and solely embeddings. We show that the incorporation of the STKG increases the prediction quality of wildfires.
{"title":"Wildfire prediction for California using and comparing Spatio-Temporal Knowledge Graphs","authors":"Martin Böckling, Heiko Paulheim, Sarah Detzler","doi":"10.1515/itit-2023-0061","DOIUrl":"https://doi.org/10.1515/itit-2023-0061","url":null,"abstract":"Abstract The frequency of wildfires increases yearly and poses a constant threat to the environment and human beings. Different factors, for example surrounding infrastructure to an area (e.g., campfire sites or power lines) contribute to the occurrence of wildfires. In this paper, we propose using a Spatio-Temporal Knowledge Graph (STKG) based on OpenStreetMap (OSM) data for modeling such infrastructure. Based on that knowledge graph, we use the RDF2vec approach to create embeddings for predicting wildfires, and we align different vector spaces generated at each temporal step by partial rotation. In an experimental study, we determine the effect of the surrounding infrastructure by comparing different data composition strategies, which involve a prediction based on tabular data, a combination of tabular data and embeddings, and solely embeddings. We show that the incorporation of the STKG increases the prediction quality of wildfires.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135192935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The use of artificial intelligence (AI) is today’s dominating technological trend across all industries. With the maturing of deep learning and other data-driven techniques, AI has over the last decade become an essential component for an increasing number of products and services. In parallel to this development, technological advances have been accelerating the production of novel AI models from large-scale datasets. This global phenomenon has been driving the need for an efficient industrialized approach to develop, manage and maintain AI models at large scale. Such an approach is provided by the state-of-the-art operational concept termed AI Factory, which refers to an infrastructure for AI models and implements the idea of AI as a Service (AIaaS). Moreover, it ensures performance, transparency and reproducibility of AI models at any point in the continuous AI development process. This concept, however, does not only require new technologies and architectures, but also new job roles. Here, we discuss current trends, outline requirements and identify success factors for AI Factories. We conclude with recommendations for their successful use in practice as well as perspectives on future developments.
{"title":"Machine learning in AI Factories – five theses for developing, managing and maintaining data-driven artificial intelligence at large scale","authors":"Wolfgang Hildesheim, Taras Holoyad, Thomas Schmid","doi":"10.1515/itit-2023-0028","DOIUrl":"https://doi.org/10.1515/itit-2023-0028","url":null,"abstract":"Abstract The use of artificial intelligence (AI) is today’s dominating technological trend across all industries. With the maturing of deep learning and other data-driven techniques, AI has over the last decade become an essential component for an increasing number of products and services. In parallel to this development, technological advances have been accelerating the production of novel AI models from large-scale datasets. This global phenomenon has been driving the need for an efficient industrialized approach to develop, manage and maintain AI models at large scale. Such an approach is provided by the state-of-the-art operational concept termed AI Factory, which refers to an infrastructure for AI models and implements the idea of AI as a Service (AIaaS). Moreover, it ensures performance, transparency and reproducibility of AI models at any point in the continuous AI development process. This concept, however, does not only require new technologies and architectures, but also new job roles. Here, we discuss current trends, outline requirements and identify success factors for AI Factories. We conclude with recommendations for their successful use in practice as well as perspectives on future developments.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135192945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract This paper explores the potential and limitations of machine learning for sensor signal identification in complex industrial systems. The objective is a tool to assist engineers in finding the correct inputs to digital twins and simulations from a set of unlabeled sensor signals. A naive end-to-end machine learning approach is usually not applicable to this task, as it would require many comparable industrial systems to learn from. We present a semi-structured approach that uses observations from the manual classification of time series and combines different algorithms to partition the set of signals into smaller groups of signals that share common characteristics. Using a real-world dataset from several power plants, we evaluate our solution for scaling-invariant measurement identification and functional relationship inference using change-point correlations.
{"title":"Machine learning in sensor identification for industrial systems","authors":"Lucas Weber, Richard Lenz","doi":"10.1515/itit-2023-0051","DOIUrl":"https://doi.org/10.1515/itit-2023-0051","url":null,"abstract":"Abstract This paper explores the potential and limitations of machine learning for sensor signal identification in complex industrial systems. The objective is a tool to assist engineers in finding the correct inputs to digital twins and simulations from a set of unlabeled sensor signals. A naive end-to-end machine learning approach is usually not applicable to this task, as it would require many comparable industrial systems to learn from. We present a semi-structured approach that uses observations from the manual classification of time series and combines different algorithms to partition the set of signals into smaller groups of signals that share common characteristics. Using a real-world dataset from several power plants, we evaluate our solution for scaling-invariant measurement identification and functional relationship inference using change-point correlations.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135044401","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Karius, Mandy Knöchel, Sascha Heße, Tim Reiprich
Abstract Cyber Security has gained a significant amount of perceived importance when talking about the risks and challenges that lie ahead in the field of information technology. A recent increase in high-profile incidents involving any form of cyber criminality have raised the awareness of threats that were formerly often hidden from public perception, e.g., with openly carried out attacks against critical infrastructure to accompany traditional forms of warfare, extending those to the cyberspace. Add to that very personal experience of everyday social engineering attacks, which are cast out like a fishing net on a large scale, e.g., to catch anyone not careful enough to double-check a suspicious email. But as the threat level rises and the attacks become even more sophisticated, so do the methods to mitigate (or at least recognize) them. Of central importance here are methods from the field of machine learning (ML). This article provides a comprehensive overview of applied ML methods in cyber security, illustrates the importance of ML for cyber security, and discusses issues and methods for generating good datasets for the training phase of ML methods used in cyber security. This includes own work on the topics of network traffic classification, the collection of real-world attacks using honeypot systems as well as the use of ML to generate artificial network traffic.
{"title":"Machine learning and cyber security","authors":"S. Karius, Mandy Knöchel, Sascha Heße, Tim Reiprich","doi":"10.1515/itit-2023-0050","DOIUrl":"https://doi.org/10.1515/itit-2023-0050","url":null,"abstract":"Abstract Cyber Security has gained a significant amount of perceived importance when talking about the risks and challenges that lie ahead in the field of information technology. A recent increase in high-profile incidents involving any form of cyber criminality have raised the awareness of threats that were formerly often hidden from public perception, e.g., with openly carried out attacks against critical infrastructure to accompany traditional forms of warfare, extending those to the cyberspace. Add to that very personal experience of everyday social engineering attacks, which are cast out like a fishing net on a large scale, e.g., to catch anyone not careful enough to double-check a suspicious email. But as the threat level rises and the attacks become even more sophisticated, so do the methods to mitigate (or at least recognize) them. Of central importance here are methods from the field of machine learning (ML). This article provides a comprehensive overview of applied ML methods in cyber security, illustrates the importance of ML for cyber security, and discusses issues and methods for generating good datasets for the training phase of ML methods used in cyber security. This includes own work on the topics of network traffic classification, the collection of real-world attacks using honeypot systems as well as the use of ML to generate artificial network traffic.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.9,"publicationDate":"2023-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48414742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hans Ole Hatzel, Haimo Stiemer, Chris Biemann, Evelyn Gius
Abstract In this article, we provide an overview of machine learning as it is applied in computational literary studies, the field of computational analysis of literary texts and literature related phenomena. We survey a number of scientific publications for the machine learning methodology the scholars used and explain concepts of machine learning and natural language processing while discussing our findings. We establish that besides transformer-based language models, researchers still make frequent use of more traditional, feature-based machine learning approaches; possible reasons for this are to be found in the challenging application of modern methods to the literature domain and in the more transparent nature of traditional approaches. We shed light on how machine learning-based approaches are integrated into a research process, which often proceeds primarily from the non-quantitative, interpretative approaches of non-digital literary studies. Finally, we conclude that the application of large language models in the computational literary studies domain may simplify the application of machine learning methodology going forward, if adequate approaches for the analysis of literary texts are found.
{"title":"Machine learning in computational literary studies","authors":"Hans Ole Hatzel, Haimo Stiemer, Chris Biemann, Evelyn Gius","doi":"10.1515/itit-2023-0041","DOIUrl":"https://doi.org/10.1515/itit-2023-0041","url":null,"abstract":"Abstract In this article, we provide an overview of machine learning as it is applied in computational literary studies, the field of computational analysis of literary texts and literature related phenomena. We survey a number of scientific publications for the machine learning methodology the scholars used and explain concepts of machine learning and natural language processing while discussing our findings. We establish that besides transformer-based language models, researchers still make frequent use of more traditional, feature-based machine learning approaches; possible reasons for this are to be found in the challenging application of modern methods to the literature domain and in the more transparent nature of traditional approaches. We shed light on how machine learning-based approaches are integrated into a research process, which often proceeds primarily from the non-quantitative, interpretative approaches of non-digital literary studies. Finally, we conclude that the application of large language models in the computational literary studies domain may simplify the application of machine learning methodology going forward, if adequate approaches for the analysis of literary texts are found.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.9,"publicationDate":"2023-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46585101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Molecular communication is a novel approach for data transmission between miniaturised devices, especially in contexts where electrical signals are to be avoided. The communication is based on sending molecules (or other particles) at nanoscale through a typically fluid channel instead of the “classical” approach of sending electrons over a wire. Molecular communication devices have a large potential in future medical applications as they offer an alternative to antenna-based transmission systems that may not be applicable due to size, temperature, or radiation constraints. The communication is achieved by transforming a digital signal into concentrations of molecules that represent the signal. These molecules are then detected at the other end of the communication channel and transformed back into a digital signal. Accurately modeling the transmission channel is often not possible which may be due to a lack of data or time-varying parameters of the channel (e.g., the movements of a person wearing a medical device). This makes the process of demodulating the signal (i.e., signal classification) very difficult. Many approaches for demodulation have been discussed in the literature with one particular approach having tremendous success – artificial neural networks. These artificial networks imitate the decision process in the human brain and are capable of reliably classifying even rather noisy input data. Training such a network relies on a large set of training data. As molecular communication as a technology is still in its early development phase, this data is not always readily available. In this paper, we discuss neural network-based demodulation approaches relying on synthetic simulation data based on theoretical channel models as well as works that base their network on actual measurements produced by a prototype test bed. In this work, we give a general overview over the field molecular communication, discuss the challenges in the demodulations process of transmitted signals, and present approaches to these challenges that are based on artificial neural networks.
{"title":"Artificial intelligence for molecular communication","authors":"Max Bartunik, J. Kirchner, Oliver Keszöcze","doi":"10.1515/itit-2023-0029","DOIUrl":"https://doi.org/10.1515/itit-2023-0029","url":null,"abstract":"Abstract Molecular communication is a novel approach for data transmission between miniaturised devices, especially in contexts where electrical signals are to be avoided. The communication is based on sending molecules (or other particles) at nanoscale through a typically fluid channel instead of the “classical” approach of sending electrons over a wire. Molecular communication devices have a large potential in future medical applications as they offer an alternative to antenna-based transmission systems that may not be applicable due to size, temperature, or radiation constraints. The communication is achieved by transforming a digital signal into concentrations of molecules that represent the signal. These molecules are then detected at the other end of the communication channel and transformed back into a digital signal. Accurately modeling the transmission channel is often not possible which may be due to a lack of data or time-varying parameters of the channel (e.g., the movements of a person wearing a medical device). This makes the process of demodulating the signal (i.e., signal classification) very difficult. Many approaches for demodulation have been discussed in the literature with one particular approach having tremendous success – artificial neural networks. These artificial networks imitate the decision process in the human brain and are capable of reliably classifying even rather noisy input data. Training such a network relies on a large set of training data. As molecular communication as a technology is still in its early development phase, this data is not always readily available. In this paper, we discuss neural network-based demodulation approaches relying on synthetic simulation data based on theoretical channel models as well as works that base their network on actual measurements produced by a prototype test bed. In this work, we give a general overview over the field molecular communication, discuss the challenges in the demodulations process of transmitted signals, and present approaches to these challenges that are based on artificial neural networks.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.9,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46443869","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
F. Maurer, Moritz Thoma, A. Surhonne, Bryan Donyanavard, A. Herkersdorf
Abstract Modern embedded and cyber-physical applications consist of critical and non-critical tasks co-located on multiprocessor systems on chip (MPSoCs). Co-location of tasks results in contention for shared resources, resulting in interference on interconnect, processing units, storage, etc. Hence, machine learning-based resource managers must operate even non-critical tasks within certain constraints to ensure proper execution of critical tasks. In this paper we demonstrate and evaluate countermeasures based on backup policies to enhance rule-based reinforcement learning to enforce constraints. Detailed experiments reveal the CPUs’ performance degradation caused by different designs, as well as their effectiveness in preventing constraint violations. Further, we exploit the interpretability of our approach to further improve the resource manager’s operation by adding designers’ experience into the rule set.
{"title":"Machine learning in run-time control of multicore processor systems","authors":"F. Maurer, Moritz Thoma, A. Surhonne, Bryan Donyanavard, A. Herkersdorf","doi":"10.1515/itit-2023-0056","DOIUrl":"https://doi.org/10.1515/itit-2023-0056","url":null,"abstract":"Abstract Modern embedded and cyber-physical applications consist of critical and non-critical tasks co-located on multiprocessor systems on chip (MPSoCs). Co-location of tasks results in contention for shared resources, resulting in interference on interconnect, processing units, storage, etc. Hence, machine learning-based resource managers must operate even non-critical tasks within certain constraints to ensure proper execution of critical tasks. In this paper we demonstrate and evaluate countermeasures based on backup policies to enhance rule-based reinforcement learning to enforce constraints. Detailed experiments reveal the CPUs’ performance degradation caused by different designs, as well as their effectiveness in preventing constraint violations. Further, we exploit the interpretability of our approach to further improve the resource manager’s operation by adding designers’ experience into the rule set.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.9,"publicationDate":"2023-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49379016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Successful and profitable companies in the 21st century have to integrate and adapt modern technologies to reach productivity and sustainability goals. While large enterprises have the resources (human, knowledge, and technology) to incorporate automation hardware and the required computational means to support automation, for small and medium enterprises (SMEs) finding the necessary resources is much more difficult. Even though benefits from such modifications are obvious (i.e., gain new clients, improve current process performance or break into new market), SMEs often lack the human resources and knowledge to implement automation. With the development of the Internet of Things (IoT), devices, sensors and platforms are becoming more affordable and available for companies and businesses, that are not connected with IT technologies, industrial sector, etc. In order to simplify automation for SMEs, simple and standardized integration procedures and best practice examples are important. In this paper we propose the concept and design of a smart bar system that is based on the Internet of Processes and Things (IoPT) concept which is able to prepare and serve drinks to clients based on smart features.
{"title":"Design of an IoPT approach to create a cocktail robot station with robotic arm: components, interfaces and control","authors":"Nataliia Klievtsova, M. Fuschlberger","doi":"10.1515/itit-2023-0006","DOIUrl":"https://doi.org/10.1515/itit-2023-0006","url":null,"abstract":"Abstract Successful and profitable companies in the 21st century have to integrate and adapt modern technologies to reach productivity and sustainability goals. While large enterprises have the resources (human, knowledge, and technology) to incorporate automation hardware and the required computational means to support automation, for small and medium enterprises (SMEs) finding the necessary resources is much more difficult. Even though benefits from such modifications are obvious (i.e., gain new clients, improve current process performance or break into new market), SMEs often lack the human resources and knowledge to implement automation. With the development of the Internet of Things (IoT), devices, sensors and platforms are becoming more affordable and available for companies and businesses, that are not connected with IT technologies, industrial sector, etc. In order to simplify automation for SMEs, simple and standardized integration procedures and best practice examples are important. In this paper we propose the concept and design of a smart bar system that is based on the Internet of Processes and Things (IoPT) concept which is able to prepare and serve drinks to clients based on smart features.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.9,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42255145","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Modern technologies such as the Internet of Things (IoT) are becoming increasingly important in various fields, including business process management (BPM) research. An important area of research in BPM is process mining, which can be used to analyze event logs e.g., to check the conformance of running processes. However, the data ingested in IoT environments often contain data quality issues (DQIs) due to system complexity and sensor heterogeneity, among other factors. To date, however, there has been little work on IoT event logs, DQIs occurring in them, and how to handle them. In this case study, we generate an IoT event log, perform a structured data quality analysis, and describe how we addressed the problems we encountered in pre-processing.
{"title":"IoT-enriched event log generation and quality analytics: a case study","authors":"J. Grüger, Lukas Malburg, Ralph Bergmann","doi":"10.1515/itit-2022-0077","DOIUrl":"https://doi.org/10.1515/itit-2022-0077","url":null,"abstract":"Abstract Modern technologies such as the Internet of Things (IoT) are becoming increasingly important in various fields, including business process management (BPM) research. An important area of research in BPM is process mining, which can be used to analyze event logs e.g., to check the conformance of running processes. However, the data ingested in IoT environments often contain data quality issues (DQIs) due to system complexity and sensor heterogeneity, among other factors. To date, however, there has been little work on IoT event logs, DQIs occurring in them, and how to handle them. In this case study, we generate an IoT event log, perform a structured data quality analysis, and describe how we addressed the problems we encountered in pre-processing.","PeriodicalId":43953,"journal":{"name":"IT-Information Technology","volume":null,"pages":null},"PeriodicalIF":0.9,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46580373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：