Search

Spatio-temporal events often describe the movements of an object in terms of space, time, and potential other attributes. Significant knowledge can be inferred by analysing them, either individually or atomically in form of trajectories. The trajectories can abstract additional properties and lead to deeper value. Moreover, external contextual information can be attributed to them to change their structure and lead to different perspectives. Because of this potentially valuable knowledge, nowadays indoor and outdoor tracking devices are used everywhere; generating countless events instantaneously. However, the extraction of knowledge from such heterogeneous, massive data is not a trivial task. In other terms, there is a need for a sophisticated system that is efficient in terms of distributed computing, failure handling, responsiveness, and abstraction. To answer this need, our study incorporates a fully fledged, reactive system for big trajectory data management. The system is unique of its kind because it is actor-based and features responsiveness, resiliency, and elasticity. Furthermore, our system is implemented using Scala; hence, we have the expressive power of both the Object-Oriented (OO) and Functional Programming (FP) paradigms. Allowing us to reach a higher level of abstraction to be able to process any trajectory type. The scope of this paper is to detail our system and discuss elasticity, routing strategies, load balancing, and our proper fault-tolerance mechanism. To fulfill this study, we consider the Geolife project’s GPS trajectory dataset.

Nowadays, distributed systems have become requisite to process and analyse the large amount of generated data. In particular, spatio-temporal data has known an exponential growth in the last years. This could be explained by the proliferation of indoor and outdoor tracking devices and the value of the knowledge that can be extracted from their analysis. The data describe a moving objects’ behavior in space and time. Once these coordinates are assembled they form a raw trajectory. The knowledge extracted by analysing the set of trajectories is very valuable; they can be even mapped to other contextual data to add value. Preprocessing is an essential step in the mining process to filter unnecessary records and clean the noise. Furthermore, reactive systems offer the possibility to process massive data in a non-blocking, resilient, responsive and elastic manner. Despite this interest, to the best of our knowledge no works have been conducted into offering a reactive system to preprocess massive trajectories. The aim of this study is to fill this gap by proposing a reactive system based on distributed actors to manage big trajectory data. Our system is deployed with the Play Framework, Akka, MongoDB, AngularJS and D3.js. Initially, the system can load batches of trajectories stored in HDFS in a distributed manner. The scope of our study is to provide an overview of the system, to study the impact of the increase in computing resources over the scalability, and to provide the optimal node configuration. The results indicate a higher scalability of the system, and the evaluation is conducted by considering the Geolife project’s GPS trajectory dataset.

Studying social influence in networks is crucial to understand how behavior spreads. An interesting number of theories were elaborated to analyze how innovations and trends get adopted. The traditional view assumes that a minority of members in a society possess qualities that make them exceptionally persuasive in spreading ideas to others. These exceptional individuals drive trends on behalf of the majority of ordinary people. They are loosely described as being informed, respected, and well connected. The leaders or influential are responsible for the dissemination of information and the propagation of influence. In this paper, we propose a new scalable and a deterministic approach for the detection of communities using leaders nodes named Leader-Community Detection Approach LCDA. The proposed approach has two main steps. The first step is the leaders’ retrieval. The second step is the community detection using similarity between nodes. Our algorithms provide good results compared to ground truth membership community.

Since the advent of social networks, Iot and Smartphones, spatial data are taken by storm by the Big Data phenomenon. Their management and analysis is a real challenge for traditional geographic information systems. Indeed, these solutions don’t respond effectively to big data constraints as they still rely on relational databases to manage and process spatial data.

Reactive Systems, especially in the Big Data Era, are undergoing a revolution. The increasing interest is due to their responsiveness, elasticity, resiliency and message-driven trait. In particular, a reactive message-driven system has one core feature asynchronous communication assuring a non-blocking behavior. A particular case of such systems is the use of actors to assure encapsulation and support fast decisions, failure handling and progess. The Actor Model meets the requirements for buiding a highly conccurent, scalable, distributed system, and Akka's tools provide the means to build it. However, to the best of our knowledge the effect of intra-actor asynchronous message processing on the routing strategies and the load balancing issue has not been addressed in the literature yet. The aim of this study is to analyse the behavior of the routers and routees over the existing routing strategies when each routee processes its messages asynchronously. Moreover, we study how the load is balanced among the routees and propose our own routing strategy that can reduce the processing time overhead by considering the complexity of the messages. For that, we build a distributed system to offline process massive trajectories, initially stored in HDFS, with Akka, the Play Framework, and MongoDB. Specifically, we leverage the Geolife projectâArZs GPS trajectory dataset.

With the increasing number of companies encompassing Big data, it is evident that the utmost challenge is to associate enormous quantity of event data to operational business processes that are extremely dynamic. To unbind the value of event data, events need to be firmly affiliated to the monitoring and control of operational processes. Nevertheless, Big data technologies converge essentially on storage, processing and rarely center on improving processes. In view of this, we advocate the integration of technologies of big data and process analysis namely Apache Kafka, Spark Streaming and PROM6. In this work, we design a scalable and distributed architecture for real time monitoring of operational business processes of a RoRo port terminal. The proposed solution permits the exploitation of process mining techniques to process large amount of events logs of several hundreds of gigabytes for process mining analysis.

The proliferation of indoor and outdoor tracking devices has led to a vast amount of spatial data. Each object can be described by several trajectories that, once analysed, can yield to significant knowledge. In particular, pattern analysis by clustering generic trajectories can give insight into objects sharing the same patterns. Still, sequential clustering approaches fail to handle large volumes of data. Hence, the necessity of distributed systems to be able to infer knowledge in a trivial time interval. In this paper, we detail an efficient, scalable and distributed execution pipeline for clustering raw trajectories. The clustering is achieved via a fuzzy similarity relation obtained by the transitive closure of a proximity relation. Moreover, the pipeline is integrated in Spark, implemented in Scala and leverages the Core and Graphx libraries making use of Resilient Distributed Datasets (RDD) and graph processing. Furthermore, a new simple, but very efficient, partitioning logic has been deployed in Spark and integrated into the execution process. The objective behind this logic is to equally distribute the load among all executors by considering the complexity of the data. In particular, resolving the load balancing issue has reduced the conventional execution time in an important manner. Evaluation and performance of the whole distributed process has been analysed by handling the Geolife project&rsquo, s GPS trajectory dataset.

Proteomic patterns can help the diagnosis of the underlying pathological state of an organ such as the ovary, the lung, and the breast, to name a few. An accurate classification of mass spectrometry is a crucial point to establish a reliable diagnosis and decision process regarding the type of cancer. A statistical methodology for classifying mass spectrometry data is proposed. An overview of wavelets, principal component analysis-T2 statistic, and support vector machines is given. The study is performed on low-mass SELDI spectra derived from patients with breast cancer and from normal controls. There are 156 samples where control (normal) patients contribute with 57 samples and 99 samples are cancer. A hyperparameter optimization is conducted to select a support vector machine classification model based on grid search. The performance was evaluated with a k-fold cross validation technique and Monte-Carlo simulation with 100 replications. The average accuracy is 100% with standard error equals to 0. The averages of the sensitivity and specificity are both equal to 100%, as well as the area under the curve. The excellent performance of our proposed method is mainly due to the statistical modeling and the feature extraction procedure proposed.

In the big data era, the cloud computing services have been adopted to face the emergence of data that needs to be stored and processed properly. However, these services need to provide safety mechanisms to insure its secure adoption. Thus, several solutions have been proposed including the use of secure architectures by customers. In that context, an architecture based on multi-agent systems has been proposed which aims to secure both storage and exploration of data hosted in the Cloud. In this paper, we present a brief synthesis of data security methods. We then focus on the multi-agent system architecture. Finally, we propose our solution considering the design and implementation in Java of an alteration agent which will ensure the secure storage of data stored in the Cloud. We finally present the test results of this agent on real datasets.

International audience; A data warehouse stores secret data about the privacy of individuals and important business activities. This makes access to this source a risk of disclosure of sensitive data. Hence the importance of implementing security measures which guarantee the data confidentiality by establishing an access control policy. In this direction, several propositions were made, but none are considered as a standard for access management to data warehouses. In this article, we will present our approach that allows first to exploit the permissions defined in the data sources in order to help the administrator to define access permissions to the data warehouse, and then our system will automatically generate the sensitivity level of each data warehouse element according to the permissions granted to an object in the data warehouse.

International audience; Respect for privacy and data confidentiality in a company are two fundamentals that must be protected. However, a Data Warehouse can be used as a very powerful mechanism for discovering crucial information, hence the importance of implementing security measures which guarantee the data confidentiality by establishing an access control policy. In this direction, several propositions were made, however, none is considered as a standard of access management to data warehouses. In this article, we will present our approach that allows first to exploit the permissions defined in the data sources in order to help the administrator to define access permissions to the data warehouse, and then our system will automatically generate the sensitivity level of each data warehouse element according to the permissions granted to an object in the data warehouse.

Recently, there has been increasing confidence for a favorable usage of big data drawn out from the huge amount of information deposited in a cloud computing system. Data kept on such systems can be retrieved through the network at the user's convenience. However, the data that users send include private information, and therefore, information leakage from these data is now a major social problem. The usage of secret sharing schemes for cloud computing have lately been approved to be relevant in which users deal out their data to several servers. Notably, in a (k, n) threshold scheme, data security is assured if and only if all through the whole life of the secret the opponent cannot compromise more than k of the n servers. In fact, a number of secret sharing algorithms have been suggested to deal with these security issues. However, a limitation of these methods is that first they do not consider long term data storage and second they assume that data tempering only occurs at retrieval time, after the distribution of the shares has been correctly done. In this paper we address these two problems by presenting a novel scheme to ensure a perpetual secure data storage and retrieval.

Complex networks have provided significant strides for understanding complex systems. One of the most relevant features of graphs representing real systems is clustering, or community structure. Those networks play an important role in the dissemination of information and influence. The hubs or influential members may be even more critical to product diffusion and the formation of widespread contagions. In this paper, we will propose a new algorithm for the community detection using influential nodes. Our algorithm uses the network topology to identify influential members, and detect community based on edge similarity named Salton similarity. Our approach doesn't require a priori knowledge of k number of communities in the network.

Cloud computing services have been adopted to provide the necessary tools and resources to face the emergence of data that needs to be stored and processed properly. However, these promising services, raise the issue of security and reliability in terms of data confidentiality, control and loss of intellectual property. In this work, we exploit the characteristics of multi agent systems to deliver an optimal and secure solution for data storage and exploration in the Cloud. Our solution is based on an encryption process before storage, while an intelligent multi agent system was designed and simulated to optimize the exploration of the data in a Cloud environment. Our architecture aims to use adaptive agents able to predict alerts, make decisions and block any intrusion.

Social networks play an important role in the dissemination of information and the spread of influence. Identifying the most influential individuals spreading information or infectious diseases can assist or hinder information dissemination, product exposure, and contagious disease detection. The leader or influential members may be even more critical to product diffusion and the formation of widespread contagions. This paper proposes a new algorithm LeadersRank, which use the network structure to identify influential members, i.e. the leaders' nodes, for community detection in social networks, without a priori knowledge of k number of communities in the network.

Mobile cloud computing has been introduced to be a powerful technology for mobile services by combining mobile computing and cloud computing technology. Though, a direct integration of two technologies can overcome a many of hurdles related to the performance, flexibility, security, and dynamic management discussed in mobile computing. Mobile cloud computing can address these problems by executing mobile applications on resource providers external to the mobile device. However, to make this vision a reality is far from being achieved and opens many new research questions. In addition, the collaboration between a mobile device and a cloud server poses complex performance issues associated with synchronization of data, network condition, security etc.

Undeniably, cloud computing has been reshaping information technology drastically, and eventually making utility computing comes true. Internet services, ranging from simple word processing to online social networking, are progressively being shifted to the cloud. These services are interesting because they assure high scalability, availability, and reliability. Nevertheless, they are missing security as a vital non-functional requirement. As users no longer control their data and hand its governance over to cloud providers, data is significantly exposed to misuses and malicious attacks. In this article we present a lightweight software based agent that runs along with the cloud database on every host machine (i.e. virtual machine). The agents implement all the logic needed to prevent malicious attacks, monitor, and assess database transactions.

Mass Spectrometry (MS) has arguably become thecore technology in proteomics. MALDI and SELDI-TOFtechniques enable the study biological fluids, e.g. human blood.Analysis of these samples can lead to discover new biomarkerswhich can ease the diagnostic and prognostic of several diseases,e.g. various cancers. In this work, we focus on MS data fromSELDI-TOF experiments. We begin with a preprocessing step inorder to remove noises due to the acquisition process of the data.Then, we apply the differential analysis to a SELDI-MS data,using the Significance Analysis of Microarray (SAM) methodimplemented in Matlab. Results using the SAM method arecompared with those obtained by the conventional t -test andAnalysis Of Variance (ANOVA) in order to evaluate its efficacyand its performance. As a result, we demonstrate that the SAMmethod can be adapted for effective significance analysis ofSELDI-MS data. It is deemed powerful and provides betterresults that totes. An easy-to-use application is developed withMatlab for mass spectrometry data analysis from raw spectra todifferential analysis, including the SAM method.