2008 IEEE/OES US/EU-Baltic International Symposium
The Gulf of Mexico hydrates research consortium has been designing a sea-floor observatory to mon... more The Gulf of Mexico hydrates research consortium has been designing a sea-floor observatory to monitor natural gas hydrates in the gulf of Mexico for almost ten years. The observatory will consist of seismo-acoustic receiving arrays, geochemical arrays in the lower water column and upper sediments as well as systems for observing microbial activity. Mississippi Canyon Lease Block 118 (MC118) in
Sleep apnea is a serious and severely under-diagnosed sleep-related respiration disorder characte... more Sleep apnea is a serious and severely under-diagnosed sleep-related respiration disorder characterized by repeated disrupted breathing events during sleep. It is diagnosed via polysomnography which is an expensive test conducted in a sleep lab requiring sleep experts to manually score the recorded data. Since the symptoms of sleep apnea are often ambiguous, it is difficult for a physician to decide whether to prescribe polysomnography. In this study, we investigate whether helpful information can be obtained by collecting and automatically analysing sleep data using a smartphone and an inexpensive strain gauge respiration belt. We evaluate how accurately we can detect sleep apnea with wide variety of machine learning techniques with data from a clinical study with 49 overnight sleep recordings. With less than one hour of training, we can distinguish between normal and apneic minutes with an accuracy, sensitivity, and specificity of 0.7609, 0.7833, and 0.7217, respectively. These res...
The presented data model is a novel approach for integrating temporal concepts into a multimedia ... more The presented data model is a novel approach for integrating temporal concepts into a multimedia database system. Multimedia objects are extended with the traditional time dimensions valid time and transaction time. In addition a new time dimension specifically tailored for multimedia data types is presented with semantics that are completely orthogonal to the already established time dimensions, valid time and transaction time, i.e., the model supports a 3D time for multimedia data. This new time dimension, the play time dimension, places the building blocks of multimedia data in a temporal structure for multimedia presentation. This model is currently being implemented in a MMDBS for distance education at UNIK, University of Oslo.
Driven by the goal to enable sleep apnea monitoring and machine learning-based detection at home ... more Driven by the goal to enable sleep apnea monitoring and machine learning-based detection at home with small mobile devices, we investigate whether interpretation-based indirect knowledge transfer can be used to create classifiers with acceptable performance. Interpretation-based indirect knowledge transfer means that a classifier (student) learns from a synthetic dataset based on the knowledge representation from an already trained Deep Network (teacher). We use activation maximization to generate visualizations and create a synthetic dataset to train the student classifier. This approach has the advantage that student classifiers can be trained without access to the original training data. With experiments we investigate the feasibility of interpretation-based indirect knowledge transfer and its limitations. The student achieves an accuracy of 97.8% on MNIST (teacher accuracy: 99.3%) with a similar smaller architecture to that of the teacher. The student classifier achieves an accu...
The Arctic Beans project is investigating the area of “Configurable and Reconfigurable Enterprise... more The Arctic Beans project is investigating the area of “Configurable and Reconfigurable Enterprise Component Architectures” . The general aim of this research is to develop more open and flexible middleware technologies, focusing on enterprise (or server-side) component architecture (such as Enterprise JavaBeans or the CORBA Component Model). The great benefit of such technologies is that they encourage re-use of components developed by third-party suppliers, and also manage distribution implicitly through the concept of a container. This works well for a significant number of application domains. However, there is increasing evidence that there are problems with this approach as the technology is applied in other areas (e.g. in mobile computing). The reason for this is that enterprise component architectures tend to enforce a certain style of distribution management in terms of, for example, the approach to persistence, security, and transactions, or the general model of interaction...
Internet services like the world-wide web and applications like News-on-Demand have become very p... more Internet services like the world-wide web and applications like News-on-Demand have become very popular over the last years. The number of users, as well as the amount of multimedia data downloaded by each user from servers in the Internet, is rapidly increasing. In this context, the potentially (very) high number of concurrent users that retrieve data from servers represents a generic problem. In the Intermediate Storage Node Concept (INSTANCE) project, we develop a radically new architecture for Media-on-Demand servers that maximizes the number of concurrent clients a single server can support. Traditional bottlenecks, like copy operations, multiple copies of the same data element in main memory, and checksum calculations in communication protocols are avoided by applying three orthogonal techniques: zero-copy-one-copy memory architecture, integrated error management, and network level framing. In this paper, we describe the design of the network level framing concept, which enabl...
Journal of Ambient Intelligence and Humanized Computing
IoT-based systems for early epidemic detection have not been investigated yet in the research. Th... more IoT-based systems for early epidemic detection have not been investigated yet in the research. The state-of-the art in sensor technology and activity recognition makes it possible to automatically detect activities of daily living (ADL). Semantic reasoning over ADLs can discover anomalies and symptoms for disorders, hence diseases and epidemics. However, semantic reasoning is computationally rather expensive and therefore unusable for real-time monitoring in large scale applications, like early epidemic detection. To overcome this limitation, this paper proposes a new scalable semantic framework based on several semantic reasoning techniques that are distributed over a semantic middleware. To reduce the number of events to process during the semantic reasoning, a complex event processing (CEP) engine is used to detect abnormal events in ADL and to generate the associated symptom indicators. To demonstrate real-time detection and scalability, the proposed framework integrates a new extension of ADLSim, a discrete event simulator that simulates long-term sequences of ADL.
Obstructive Sleep Apnea (OSA) is a common, but severely under-diagnosed sleep disorder that affec... more Obstructive Sleep Apnea (OSA) is a common, but severely under-diagnosed sleep disorder that affects the natural breathing cycle during sleep with periods of reduced respiration or no airflow at all. It is our long-term goal to increase the percentage of diagnosed OSA cases and reduce the time to diagnosis with user friendly and cost-efficient tools for sleep analysis at home. As a first step towards this goal, we study in this work whether a small subset of those physiological signals that are used in classical OSA diagnosis in combination with automatic classification allow to detect apnea events. We study the performance of five data mining techniques to classify epochs of data from the Apnea-ECG and MIT-BIH databases from PhysioNet as either disrupted or normal breathing. The data is only slightly preprocessed (rate reduction and normalization). We focus in this study on respiratory signals from the nose, abdomen, chest, and oxygen saturation. We measure for any combination of these signals the accuracy, sensitivity, specificity and Kappa statistics of classification with Artificial Neural Network, Support Vector Machine, Decision Tree, K-Nearest Neighbor and Random Forest. For Apnea-ECG, we achieve an accuracy of 96.6% with a combination of respiration data from the chest and nose as input data; and an accuracy of more than 90% for all signal combinations. Interestingly, these good results are also achieved with the simple KNN technique. The results for MIT-BIH are lower, because of noise, smaller size, and some class imbalance. The accuracy does not significantly improve with the number of signals included in the signal combinations. We conclude that one signal might be sufficient to detect disrupted breathing, if the data set is of sufficient quality and size, and that respiration from the abdomen is the preferrable choice when taking into account both classification performance and patient comfort.
In network monitoring, an important issue is the number of tuples the data stream management syst... more In network monitoring, an important issue is the number of tuples the data stream management system (DSMS) can handle for different network loads. In order to gracefully handle overload situations, some DSMSs are equipped with a tuple dropping functionality, also known as load shedding. These DSMSs register and relate the number of received and dropped tuples, i.e., the relative throughput, and perform different kinds of calculations on them. Over the past few years, several solutions and methods have been suggested to efficiently perform load shedding. The simplest approach is to keep a count of all the dropped tuples, and to report this to the end user. In our experiments, we study two DSMSs, i.e., TelegraphCQ with support for load shedding, and STREAM without this support. We use three particular network monitoring tasks to evaluate the two DSMS with respect to their ability of load shedding and performance. We demonstrate that it is important to investigate the correctness of load shedding by showing that the reported number of dropped tuples is not always correct.
Lately, on-demand streaming multimedia applications have become very popular. Contemporary person... more Lately, on-demand streaming multimedia applications have become very popular. Contemporary personal computers can handle the load imposed by such multimedia applications on the client side, but the potentially high number of concurrent users accessing a server represents a generic problem. The multimedia storage system is responsible for storage and retrieval of multimedia data from storage devices, and plays a vital role for the performance and scalability of multimedia servers. It deals with issues related to data placement, scheduling, file management, continuous data delivery, buffer management, prefetching, etc., and with the particular demands of multimedia applications, such as real-time characteristics, large file sizes, high data rates, and several data sources. Performing these tasks and supporting these requirements appropriately are burdened by an increasing speed mismatch between processors and the most prolific and affordable storage devices,-magnetic disks-, and by the introduction of new requirements in new multimedia scenarios. In this article, we give a survey of storage system support for continuous media applications and discuss issues related to integration of different mechanisms for the future multimedia storage systems.
Efficient information sharing is very important for emergency and rescue operations. These operat... more Efficient information sharing is very important for emergency and rescue operations. These operations often have to be performed in environments where no communication infrastructure exists. Therefore, Mobile Ad-Hoc Network (MANET) technologies need to be leveraged for generic information sharing services. However, MANETs are using the shared medium of wireless networks, i.e., IEEE 802.11, which means that bandwidth is limited and mobility might cause network partitions. In order to achieve high availability of important information and increase overall system performance, replication can be used. The drawback of aggressive replication is the problem of inconsistencies and the costs of possible conflict resolution strategies. Since emergency and rescue operations are only operating for a short time compared to classical database applications, storage space might not be considered a bottleneck even for small devices: Instead of deleting and updating data, data versions are created, and each application can influence the choice of consistency model for its data and use its own policy to resolve conflicts. Such an approach does not only allow for application specific conflict resolution, but also enables traceability of what happened and has been registered, in the aftermath of the emergency and rescue operation.
Mobile Ad-Hoc Networks maintain information in the routing table about reachable nodes. In emerge... more Mobile Ad-Hoc Networks maintain information in the routing table about reachable nodes. In emergency and rescue operations, human groups play an important role. This is visible at the network level as independent network partitions which are for some time stable before their members change through merging or partitioning. We use the information from stable routing tables to optimize the synchronization
Proceedings of the 12th International Conference on Telecommunications, 2013
ABSTRACT The dynamically changing set of multimedia capable devices in the vicinity of a user can... more ABSTRACT The dynamically changing set of multimedia capable devices in the vicinity of a user can be leveraged to create new ways of experiencing multimedia applications through migrating parts of running multimedia applications to the most suited devices. This paper addresses one of the core challenges of application migration, i.e., migration of transport protocol state that is maintained by the endpoints of established connections. Our solution fulfills the stringent temporal requirements of multimedia applications and enables migratable applications to interact with legacy applications, e.g., a migratable video player together with YouTube. The core idea of our solution, called SOCKMAN, is to provide a middleware service to hide that proxy-based forwarding is used to migrate connection endpoints, i.e. sockets, and to maintain an end-to-end perspective for the applications. The evaluation of the SOCKMAN implementation shows that SOCKMAN meets multimedia application requirements, preserves transport protocol state, and performs well on low-end devices, like mobile phones.
The trends over the recent years show an increase in stream- ing media transported across the Int... more The trends over the recent years show an increase in stream- ing media transported across the Internet in addition to al- ready abundant web traffic, increasing the need for distri- bution schemes that efficiently handle this type of traffic. There are two obvious bottlenecks in any content distribu- tion system, namely the origin server and the network. Our goal is
Proceedings of the 8th International Conference on Telecommunications, 2005. ConTEL 2005., 2005
Summary form only given. A new class of data-intensive applications, such as sensor networks, net... more Summary form only given. A new class of data-intensive applications, such as sensor networks, network traffic analysis, financial tickers, Web or telecommunications transaction log analysis, has become widely recognized. These applications require support for on-line analysis of rapidly changing data streams. However, traditional database management systems (DBMS) have no pre-defined notion of time and cannot handle data on-line (i.e., in
Abstract Peer-to-peer (P2P) based networks have several desirable features for content distributi... more Abstract Peer-to-peer (P2P) based networks have several desirable features for content distribution, such as low costs, scalability, and fault tolerance. However, they fail to provide guarantees for content delivery. In order to combine the desired features of classical Content ...
2008 IEEE/OES US/EU-Baltic International Symposium
The Gulf of Mexico hydrates research consortium has been designing a sea-floor observatory to mon... more The Gulf of Mexico hydrates research consortium has been designing a sea-floor observatory to monitor natural gas hydrates in the gulf of Mexico for almost ten years. The observatory will consist of seismo-acoustic receiving arrays, geochemical arrays in the lower water column and upper sediments as well as systems for observing microbial activity. Mississippi Canyon Lease Block 118 (MC118) in
Sleep apnea is a serious and severely under-diagnosed sleep-related respiration disorder characte... more Sleep apnea is a serious and severely under-diagnosed sleep-related respiration disorder characterized by repeated disrupted breathing events during sleep. It is diagnosed via polysomnography which is an expensive test conducted in a sleep lab requiring sleep experts to manually score the recorded data. Since the symptoms of sleep apnea are often ambiguous, it is difficult for a physician to decide whether to prescribe polysomnography. In this study, we investigate whether helpful information can be obtained by collecting and automatically analysing sleep data using a smartphone and an inexpensive strain gauge respiration belt. We evaluate how accurately we can detect sleep apnea with wide variety of machine learning techniques with data from a clinical study with 49 overnight sleep recordings. With less than one hour of training, we can distinguish between normal and apneic minutes with an accuracy, sensitivity, and specificity of 0.7609, 0.7833, and 0.7217, respectively. These res...
The presented data model is a novel approach for integrating temporal concepts into a multimedia ... more The presented data model is a novel approach for integrating temporal concepts into a multimedia database system. Multimedia objects are extended with the traditional time dimensions valid time and transaction time. In addition a new time dimension specifically tailored for multimedia data types is presented with semantics that are completely orthogonal to the already established time dimensions, valid time and transaction time, i.e., the model supports a 3D time for multimedia data. This new time dimension, the play time dimension, places the building blocks of multimedia data in a temporal structure for multimedia presentation. This model is currently being implemented in a MMDBS for distance education at UNIK, University of Oslo.
Driven by the goal to enable sleep apnea monitoring and machine learning-based detection at home ... more Driven by the goal to enable sleep apnea monitoring and machine learning-based detection at home with small mobile devices, we investigate whether interpretation-based indirect knowledge transfer can be used to create classifiers with acceptable performance. Interpretation-based indirect knowledge transfer means that a classifier (student) learns from a synthetic dataset based on the knowledge representation from an already trained Deep Network (teacher). We use activation maximization to generate visualizations and create a synthetic dataset to train the student classifier. This approach has the advantage that student classifiers can be trained without access to the original training data. With experiments we investigate the feasibility of interpretation-based indirect knowledge transfer and its limitations. The student achieves an accuracy of 97.8% on MNIST (teacher accuracy: 99.3%) with a similar smaller architecture to that of the teacher. The student classifier achieves an accu...
The Arctic Beans project is investigating the area of “Configurable and Reconfigurable Enterprise... more The Arctic Beans project is investigating the area of “Configurable and Reconfigurable Enterprise Component Architectures” . The general aim of this research is to develop more open and flexible middleware technologies, focusing on enterprise (or server-side) component architecture (such as Enterprise JavaBeans or the CORBA Component Model). The great benefit of such technologies is that they encourage re-use of components developed by third-party suppliers, and also manage distribution implicitly through the concept of a container. This works well for a significant number of application domains. However, there is increasing evidence that there are problems with this approach as the technology is applied in other areas (e.g. in mobile computing). The reason for this is that enterprise component architectures tend to enforce a certain style of distribution management in terms of, for example, the approach to persistence, security, and transactions, or the general model of interaction...
Internet services like the world-wide web and applications like News-on-Demand have become very p... more Internet services like the world-wide web and applications like News-on-Demand have become very popular over the last years. The number of users, as well as the amount of multimedia data downloaded by each user from servers in the Internet, is rapidly increasing. In this context, the potentially (very) high number of concurrent users that retrieve data from servers represents a generic problem. In the Intermediate Storage Node Concept (INSTANCE) project, we develop a radically new architecture for Media-on-Demand servers that maximizes the number of concurrent clients a single server can support. Traditional bottlenecks, like copy operations, multiple copies of the same data element in main memory, and checksum calculations in communication protocols are avoided by applying three orthogonal techniques: zero-copy-one-copy memory architecture, integrated error management, and network level framing. In this paper, we describe the design of the network level framing concept, which enabl...
Journal of Ambient Intelligence and Humanized Computing
IoT-based systems for early epidemic detection have not been investigated yet in the research. Th... more IoT-based systems for early epidemic detection have not been investigated yet in the research. The state-of-the art in sensor technology and activity recognition makes it possible to automatically detect activities of daily living (ADL). Semantic reasoning over ADLs can discover anomalies and symptoms for disorders, hence diseases and epidemics. However, semantic reasoning is computationally rather expensive and therefore unusable for real-time monitoring in large scale applications, like early epidemic detection. To overcome this limitation, this paper proposes a new scalable semantic framework based on several semantic reasoning techniques that are distributed over a semantic middleware. To reduce the number of events to process during the semantic reasoning, a complex event processing (CEP) engine is used to detect abnormal events in ADL and to generate the associated symptom indicators. To demonstrate real-time detection and scalability, the proposed framework integrates a new extension of ADLSim, a discrete event simulator that simulates long-term sequences of ADL.
Obstructive Sleep Apnea (OSA) is a common, but severely under-diagnosed sleep disorder that affec... more Obstructive Sleep Apnea (OSA) is a common, but severely under-diagnosed sleep disorder that affects the natural breathing cycle during sleep with periods of reduced respiration or no airflow at all. It is our long-term goal to increase the percentage of diagnosed OSA cases and reduce the time to diagnosis with user friendly and cost-efficient tools for sleep analysis at home. As a first step towards this goal, we study in this work whether a small subset of those physiological signals that are used in classical OSA diagnosis in combination with automatic classification allow to detect apnea events. We study the performance of five data mining techniques to classify epochs of data from the Apnea-ECG and MIT-BIH databases from PhysioNet as either disrupted or normal breathing. The data is only slightly preprocessed (rate reduction and normalization). We focus in this study on respiratory signals from the nose, abdomen, chest, and oxygen saturation. We measure for any combination of these signals the accuracy, sensitivity, specificity and Kappa statistics of classification with Artificial Neural Network, Support Vector Machine, Decision Tree, K-Nearest Neighbor and Random Forest. For Apnea-ECG, we achieve an accuracy of 96.6% with a combination of respiration data from the chest and nose as input data; and an accuracy of more than 90% for all signal combinations. Interestingly, these good results are also achieved with the simple KNN technique. The results for MIT-BIH are lower, because of noise, smaller size, and some class imbalance. The accuracy does not significantly improve with the number of signals included in the signal combinations. We conclude that one signal might be sufficient to detect disrupted breathing, if the data set is of sufficient quality and size, and that respiration from the abdomen is the preferrable choice when taking into account both classification performance and patient comfort.
In network monitoring, an important issue is the number of tuples the data stream management syst... more In network monitoring, an important issue is the number of tuples the data stream management system (DSMS) can handle for different network loads. In order to gracefully handle overload situations, some DSMSs are equipped with a tuple dropping functionality, also known as load shedding. These DSMSs register and relate the number of received and dropped tuples, i.e., the relative throughput, and perform different kinds of calculations on them. Over the past few years, several solutions and methods have been suggested to efficiently perform load shedding. The simplest approach is to keep a count of all the dropped tuples, and to report this to the end user. In our experiments, we study two DSMSs, i.e., TelegraphCQ with support for load shedding, and STREAM without this support. We use three particular network monitoring tasks to evaluate the two DSMS with respect to their ability of load shedding and performance. We demonstrate that it is important to investigate the correctness of load shedding by showing that the reported number of dropped tuples is not always correct.
Lately, on-demand streaming multimedia applications have become very popular. Contemporary person... more Lately, on-demand streaming multimedia applications have become very popular. Contemporary personal computers can handle the load imposed by such multimedia applications on the client side, but the potentially high number of concurrent users accessing a server represents a generic problem. The multimedia storage system is responsible for storage and retrieval of multimedia data from storage devices, and plays a vital role for the performance and scalability of multimedia servers. It deals with issues related to data placement, scheduling, file management, continuous data delivery, buffer management, prefetching, etc., and with the particular demands of multimedia applications, such as real-time characteristics, large file sizes, high data rates, and several data sources. Performing these tasks and supporting these requirements appropriately are burdened by an increasing speed mismatch between processors and the most prolific and affordable storage devices,-magnetic disks-, and by the introduction of new requirements in new multimedia scenarios. In this article, we give a survey of storage system support for continuous media applications and discuss issues related to integration of different mechanisms for the future multimedia storage systems.
Efficient information sharing is very important for emergency and rescue operations. These operat... more Efficient information sharing is very important for emergency and rescue operations. These operations often have to be performed in environments where no communication infrastructure exists. Therefore, Mobile Ad-Hoc Network (MANET) technologies need to be leveraged for generic information sharing services. However, MANETs are using the shared medium of wireless networks, i.e., IEEE 802.11, which means that bandwidth is limited and mobility might cause network partitions. In order to achieve high availability of important information and increase overall system performance, replication can be used. The drawback of aggressive replication is the problem of inconsistencies and the costs of possible conflict resolution strategies. Since emergency and rescue operations are only operating for a short time compared to classical database applications, storage space might not be considered a bottleneck even for small devices: Instead of deleting and updating data, data versions are created, and each application can influence the choice of consistency model for its data and use its own policy to resolve conflicts. Such an approach does not only allow for application specific conflict resolution, but also enables traceability of what happened and has been registered, in the aftermath of the emergency and rescue operation.
Mobile Ad-Hoc Networks maintain information in the routing table about reachable nodes. In emerge... more Mobile Ad-Hoc Networks maintain information in the routing table about reachable nodes. In emergency and rescue operations, human groups play an important role. This is visible at the network level as independent network partitions which are for some time stable before their members change through merging or partitioning. We use the information from stable routing tables to optimize the synchronization
Proceedings of the 12th International Conference on Telecommunications, 2013
ABSTRACT The dynamically changing set of multimedia capable devices in the vicinity of a user can... more ABSTRACT The dynamically changing set of multimedia capable devices in the vicinity of a user can be leveraged to create new ways of experiencing multimedia applications through migrating parts of running multimedia applications to the most suited devices. This paper addresses one of the core challenges of application migration, i.e., migration of transport protocol state that is maintained by the endpoints of established connections. Our solution fulfills the stringent temporal requirements of multimedia applications and enables migratable applications to interact with legacy applications, e.g., a migratable video player together with YouTube. The core idea of our solution, called SOCKMAN, is to provide a middleware service to hide that proxy-based forwarding is used to migrate connection endpoints, i.e. sockets, and to maintain an end-to-end perspective for the applications. The evaluation of the SOCKMAN implementation shows that SOCKMAN meets multimedia application requirements, preserves transport protocol state, and performs well on low-end devices, like mobile phones.
The trends over the recent years show an increase in stream- ing media transported across the Int... more The trends over the recent years show an increase in stream- ing media transported across the Internet in addition to al- ready abundant web traffic, increasing the need for distri- bution schemes that efficiently handle this type of traffic. There are two obvious bottlenecks in any content distribu- tion system, namely the origin server and the network. Our goal is
Proceedings of the 8th International Conference on Telecommunications, 2005. ConTEL 2005., 2005
Summary form only given. A new class of data-intensive applications, such as sensor networks, net... more Summary form only given. A new class of data-intensive applications, such as sensor networks, network traffic analysis, financial tickers, Web or telecommunications transaction log analysis, has become widely recognized. These applications require support for on-line analysis of rapidly changing data streams. However, traditional database management systems (DBMS) have no pre-defined notion of time and cannot handle data on-line (i.e., in
Abstract Peer-to-peer (P2P) based networks have several desirable features for content distributi... more Abstract Peer-to-peer (P2P) based networks have several desirable features for content distribution, such as low costs, scalability, and fault tolerance. However, they fail to provide guarantees for content delivery. In order to combine the desired features of classical Content ...
Uploads
Papers by Vera Goebel