Search

Nowadays, tremendous growth of Internet of Things (IoT) applications is seen in smart environments such as medical remote care applications which are crucial due to the general aging of the population. With the recent advancements in IoT-based healthcare technologies, activity recognition can be used as the key part of the intelligent healthcare systems to monitor elderly people to live independently at homes and promote a better care. Recently, the evidence theory and its derivates approaches began to take place in the fields of activity recognition in these smart systems. However, these approaches are generally inconsistent with the probability calculus due to the lower and upper probability bounds considering the combined evidences. To overcome these challenges and to get more precisely the reconcilement between the evidence theory with the frequentist approach of probability calculus, this work proposes a new methodology for combining beliefs, addressing some of the disadvantages exhibited by the evidence theory and its derivatives. This methodology merges the non-normalized conjunctive and the majority rules. The proposed rule is evaluated in numerical simulation case studies involving the activity recognition in a smart home environment. The results show that this strategy produces intuitive results in favor of the more committed hypothesis.

A number of conventional clustering algorithms suffer from poor scalability, especially for data lake. Thus many modified clustering algorithms have been proposed to speed up these conventional algorithms based on the employment of data sampling techniques. However, these representations require the number of clusters to proceed to centroid selection for final data clustering. To address this limitation, this paper develops a two-phase clustering-based methodology. In the first phase, rather than attempting to construct a random sampling, we define a novel approach that computes plausible sample points, uses them as centroids for the final clusters. To speedup our clustering algorithm in the second phase we propose a parallelization scheme in conjunction with a Spark parallel processing implementation. Computational experiments reveal that the Global sampling method is more effective in terms of both quality and stability compared to the most popular K-means algorithm for the same parameter settings.

International audience; Building context-aware pervasive computing systems - such as ambient intelligent spaces or ubiquitous robots - needs to take into account the quality of contextual information collected from sensors. Such information are often inaccurate, uncertain or subject to noise due to environment and user dynamics. Dempster-Shafer theory has been extensively adopted to handle uncertainty in situation and activity recognition. This theory is used to represent, manipulate and decide under uncertainty. However, combining information using Dempster's rule may produce counterintuitive decision in highly conflicting evidences due to sources failure. Recently, a variety of rules were proposed to overcome such drawback. Inspired by Murphy's rule, we propose in this paper a new rule called “Weighted Average Combination Rule” (WACR) to deal with context recognition in highly dynamic environment such as ambient intelligence spaces. The proposed WACR rule is based on evidence arithmetic average and cardinality. WACR rule was applied to some conflictual evidence examples and has been shown to reap more appropriate decisions than other alternative rules for decision-making in activity-aware systems. To demonstrate the applicability and performance of our approach, we have studied a scenario of context recognition in an ambient intelligent environment. In this scenario, we simulated a smart kitchen composed of status devices and RFID sensors that allow determining what is the artifact in use by the inhabitant and for which activity.

International audience; In recent years, a plethora of different studies for design of traditional ensemble classifiers has been proposed in order to improve final recognition accuracy. However, among the ensemble classifiers, combination methods are focused on building independent classifiers of the same or different algorithms using majority voting methods. In this paper, we present a new fusion scheme for ensemble classifiers based on a new concept called Generalized Fuzzy Soft Set (GFSS), which we apply in activity classification. Essentially, we apply a weighted aggregate operator to the output of each classifier in order to fuse the GFSS into a more reliable classifier. The proposed fusion method is based on a new ranking algorithm to classify activities. We show that the proposed method produces more accurate results than the best single classifier and its effectiveness is demonstrated by comparing it with single classifier in terms of activity recognition accuracy.

The present work attempts to build a bio-cryptographic system that combines transformed minutiae pairwise feature and user-generated password fuzzy vault. The fingerprint fuzzy vault is based on a new minutiae pairwise structure, which overcomes the fingerprint feature publication while the secret binary vault code is generated according to the fingerprint fuzzy vault result. The authentication process involves two stages: fuzzy vault matching and secret vault code validation. Our minutiae pairwise transformation produces different templates thus resolving the problem of cross matching attacks in fingerprint fuzzy vault. So, the original fingerprint template cannot be recreated because it is protected by the key generated from the user password. In addition, the proposed bio-cryptographic system ensures an acceptable security level for user authentication.

With the growing emergence of ambient intelligence, ubiquitous computing, sensor networks and wireless networking technologies, ''ubiquitous networked robotics'' is becoming an active research domain of intelligent autonomous systems. It targets new innovative applications in which robotic systems will become part of these networks of artifacts to provide novel capabilities and various assistive services anywhere and anytime, such as healthcare and monitoring services for elderly in Ambient Assisted Living (AAL) environments. Situation recognition, in general, and activity recognition, in particular, provide an added value on the contextual information that can help the ubiquitous networked robot to autonomously provide the best service that meet the needs of the elderly. Dempster-Shafer theory of evidence and its derivatives are an efficient tool to handle uncertainty and incompleteness in smart homes and ubiquitous computing environments. However, their combination rules yield counter-intuitive results in high conflicting activities. In this paper, we propose a new approach to support conflict resolution in activity recognition in AAL environments. This approach is based on a new mapping for conflict evidential fusion to increase the efficiency and accuracy of activity recognition. It gives intuitive interpretation for combining multiple sources in all conflicting situations. The proposed approach, evaluated on a real world smart home dataset, achieves 78% of accuracy in activity recognition. The obtained results outperform those obtained with the existing combination rules.

International audience; The metacomputing environments are becoming real distributed running platforms for compute-intensive services. One of the most difficult problems to be solved by metacomputing systems is ensuring accurate and fast prediction of available performance on each resource. The main objective of the present study is to develop a new prediction model that can be used to predict the future CPU load in a distributed computing environment. This prediction model is based on a mixture of Adaptive Network based Fuzzy Inference Systems (ANFIS) via the naïve Bayes assumption. Experimental results for different load time series confirm that the new prediction model performs better than other CPU load prediction methods. In addition, a comparison with previous prediction methods to evaluate their accuracy is presented.

The increasing amount of available XML documents collections has led to the emergence of new challenges in information retrieval field. Therefore, multiple sources of evidence were used to retrieve XML elements at different levels of granularity. XML information retrieval combines textual and structural information to perform different information retrieval tasks. In this paper, we propose a new approach exploiting link evidence to re-rank XML retrieval results. Our approach, based on fuzzy logic concepts, combines both content and link evidence for all retrieved XML elements. The combination process generates a new ranked list from the initial returned list. Experiment based on INEX 2007 Wikipedia collection showed improvement of the interpolated precision values.

Activity recognition is an important task which can be applied to many real-life problems in pervasive computing. In this work, we propose concurrent activities recognition system based on two layers of inference. The first layer develops a framework for dealing with fusion system through merging different sources of information using evidence theory. The second layer proposes a decision framework under the fuzzy logic formalism. Our experimental results suggest that the fuzzy logic method for the plausibility combinations at the decision level is the best for activities in progress simultaneously but not necessarily involving the user's interaction at the same time steps. It yields high accuracy of 79.7%, regarding experimental results.

In order to have a normal behavior in combinationof bodies of evidence, this paper proposes a new combinationrule. This rule includes the cardinality of focal set elements inconjunctive operation and the conflict redistribution to all steps.Based on the focal set cardinalities, the conflict redistributionis assigned using factors. The weighted factors are computedfrom the original and the conjunctive masses assigned to eachfocal element. This strategy forces the conflict redistribution infavor of the more committed hypothesis. Our method is evaluatedand compared with some numerical examples reported in theliterature. As result, this rule redistributes the conflict in favor ofthe more committed hypothesis and gives intuitive interpretationfor combining multiple information sources with coherent results. &nbsp

Scientific applications are very complex and need massive computing power and storage space. Distributed computing environment has become a new technology to execute large-scale applications and Cloud computing is one of these technologies. Resource allocation is one of the most important challenges in the Cloud Computing. The optimally assigning of the available resources to the needed cloud applications is known to be a NP complete problem. In this paper, we propose a new task scheduling strategy for resource allocation for maximizing profit in cloud computing environment. We focus on minimizing the total executing time (makespan) of task scheduling and maximizing the resources exploitation. To show the interest of the proposed solution, Experiments results are conducted on a simulation data set.

Query Expansion is an important component for information retrieval systems. It makes possible the reformulation of the initial user query by adding new terms. In this paper, we propose a new approach for term selection in the relevance feedback process. This approach, based on Rocchio formula, is an adaptation to the XML information retrieval context. It can resolve two major problems specific to the XML information retrieval: the overlapping problem in the list of retrieved elements; and the problem of inclusion of irrelevant elements in the selection of new query terms.

Distributed computing environment has become a new technology to execute large-scale applications and Cloud computing is one of these technologies. Resource allocation is one of the most important challenges in the Cloud Computing. The optimally assigning of the available resources to the needed cloud applications is known to be a NP complete problem. In this paper, we propose a new task scheduling strategy based on the total order for resource allocation to improve the Min-Min algorithm. We focus on minimizing the total executing time (makespan) of task scheduling and maximizing the use of resources. Experimental results demonstrate that the proposed approach permits more adaptive resources allocation for independent jobs scheduling in the cloud computing environment.

We propose in this paper a new evidential link-based approach for re-ranking XML retrieval results by using the Dempster-Shafer theory of evidence to combine content relevance evidence for each retrieved XML element with its computed link evidence (score and rank). The use of the Dempster–Shafer theory is motivated by the need to improve retrieval accuracy by incorporating the uncertain nature of both bodies of evidence (content and link relevance). The link score is computed according to a new link analysis algorithm based on weighted links, where relevance is propagated through the two types of links, i.e., hierarchical and navigational. The propagation, i.e. the amount of relevance score received by each retrieved XML element, depends on link weight which is defined according to two parameters: link type and link length. To evaluate our proposal we carried out a set of experiments based on INEX data collection.

International audience; These last years, a lot of combination rules emerged in order to model the situations of belief fusion. These rules can be classified in two different classes. However, these rules do not differentiate between focal elements in the combination step which produce counterintuitive results in some situations. Motivated by this observation, we propose a new combination rule which hybrids the strategies of these two classes. Our rule is two-step operator where the averaging step comes first, and then the conflict redistribution step. Experimental studies are conducted on a real smart home dataset to show the accuracy of our rule in ubiquitous-assisted living situation.

In ambient intelligence environments, the information provided by robot's embedded sensors and physical or logical entities may be inaccurate and uncertain. The Dempster-Shafer evidence Theory (DST) gives a mathematical convenient framework for the evidential fusion representation and inference of uncertain information. However, DST yields counterintuitive results in high conflicting ambient intelligence situations. This paper aims to provide a new strategy to manage conflict in activity recognition process in the ambient intelligence applications. It addresses the challenge of uncertainty and proposes an evidential fusion model based on the management of conflicting situation to optimize decision making in activity recognition. The proposed approach gives intuitive interpretation for combining multiple sources in conflicting situations and avoids the problems of using The Dempster-Shafer rule of combination.

International audience; Ensuring adequate use of the computing resources for highly fluctuating availability in multi-user computational environments requires effective prediction models, which play a key role in achieving application performance for large-scale distributed applications. Predicting the processor availability for scheduling a new process or task in a distributed environment is a basic problem that arises in many important contexts. The present paper aims at developing a model for single-step-ahead CPU load prediction that can be used to predict the future CPU load in a dynamic environment. Our prediction model is based on the control of multiple Local Adaptive Network-based Fuzzy Inference Systems Predictors (LAPs) via the Naïve Bayesian Network inference between clusters states of CPU load time points obtained by the C-means clustering process. Experimental results show that our model performs better and has less overhead than other approaches reported in the literature.

Tasks scheduling in heterogeneous computing environments is one of the most challenging problems in distributed computing. The optimally mapping of independent tasks onto heterogeneous distributed computing systems is known to be NP complete problem. This paper addresses a two-stage methodology for solving the independent task scheduling problems in heterogeneous distributed computing. The scheduler aims to minimize the total completion time using the task reassignment strategy. This later uses a new Makespan Refinery Approach (MRA) to improve our initial task scheduling solution by reducing the maximum completion time. The effectiveness of the proposed scheduling method has been tested and evaluated using simulations. The experiment results show the behaviour of the scheduling method for the short completion time of a set of tasks.

International audience; Resources performance forecasting constitutes one of particularly significant research problems in distributed computing. To ensure an adequate use of the computing resources in a metacomputing environment, there is a need for effective and flexible forecasting method to determine the available performance on each resource. In this paper, we present a modeling approach to estimating the future value of CPU load. This modeling prediction approach uses the combination of adaptive network-based fuzzy inference systems (ANFIS) and the clustering process applied on the CPU Load time series. Experiments show the feasibility and effectiveness of this approach that achieves significant improvement and outperforms the existing CPU load prediction models reported in literature.

National audience; Fingerprint matching is one of the most important problems in an Automatic Fingerprint Identification System. The Fingerprint matching is a high computational task, repetitive and arduous. Moreover, fingerprint databases are by nature large scale. Indeed, there is a need for flexible and independent solution for automatic distribution of fingerprint matching. So, this paper introduces an agent-based distributed matching system. The matching task distributed on a available local Network resources with an optimal way in order to ensure a fast matching process in large scale fingerprint databases. The distribution workload is ensured by our developed middleware allowing the monitoring and prediction of unexploited network computing resources, and designed for the implementation of intensive calculation services.

This paper presents a new fingerprint matching algorithm based on inter-minutiae relational network. This latter is represented by constraint graph (binary CSP graph) which uses, in addition to distances between minutiae, ridge counts and the relative ridge count angles according to the both minutiae pair orientations. Therefore, the resulting minutiae features are invariant to the geometric transformation and the matching process does not require an implicit alignment of the two fingerprint images. The main advantage of our approach is that it can be used for partial fingerprints. Finally, we report the results of testing our method on the public fingerprint databases DB1 and DB2 FVC2002.

In this paper, we propose an original fingerprint matching algorithm using a set of intrinsic coordinate systems which each one is attached to each minutia according to its orientation estimated from fingerprint image. Exploiting these coordinate systems, minutiae locations can be redefined by means of projection of these minutiae coordinates on the relative reference of each orientation minutia. Thus, our matching algorithm use these relative minutiae coordinate to calculate the matching score between two fingerprints. To avoid the directional field estimation errors, we propose a minutia orientation variation in order to manage the projection errors of the location minutiae. With this technique, our approach doesn't embedding fingerprint alignment into the minutia matching stage to design the robust algorithms for fingerprint recognition. The algorithm matching was tested on a fingerprint database DB2 used in FVC2000 and the results are promising.

The purpose of this paper is to investigate a new methodology for default reasoning in presence of priorities which are represented by an arbitrary partial order on default rules. In a previous work, we have defined an alternative approach for characterizing extensions in prioritized default theories initially proposed by Brewka. We present here another approach for computing default proofs for query answering in prioritized normal default logic.