Computational Intelligence and Its Applications

Choices of decision makers in a basketball team are not limited to the strategies to be adopted during games. The most important ones are outside the field and concern team composition and talented and productive players to acquire on which the team can rely to raise its game level. In this paper, we propose to use data mining tasks to help decision makers to make appropriate decisions that will lead to the improvement of the performance of their players and their team. Tasks such as clustering, classification and regression are used to detect weaknesses of a team; best players that can help overcome these weaknesses; predict performance and salaries of players. These will be done on the NBA dataset.

The spatial data mining (SDM) is a process that extracts knowledge from large volumes of spatial data. It takes into account the spatial relationships between the data. To integrate these relations in the mining process, SDM uses two main approaches: Static approach that integrates spatial relationships in a preprocessing phase, and dynamic approach that takes into consideration the spatial relationship during the process. In this work, we are interested in this last approach. Our proposition consists on taking into consideration the spatial component in the similarity measure. We propose two similarity measures; $$d_{Dyn1}$$, $$d_{Dyn2}$$. We will use those distances on the main task of SDM, spatial clustering, particularly on K-means algorithm. Moreover, a comparaison between these two approaches and other methods of clustering will be given. The tests are conducted on Boston dataset with 590 objects.

In spite of the efforts made in the Arabic language on the syntactic and semantic level, it remains very restricted, even those on the Arabic Sacred Book are few and very limited, due to its difficulties and peculiarities. In this paper we tried to shed the light on some of the recent works that have been conducted to present a semantic representation and manipulation of the Islamic texts to define the problems, limitations and the possible future works that need our intention to improve the semantic support in the Arabic religious texts. Furthermore, we intent to briefly present our project that aims to help us reading, understanding, and interpreting the Islamic legislative sources. The goal of this project is divided into two main tasks which are the creation of an ontology representing the Islamic knowledge and the development of a system which can analyze this knowledge. The ultimate goal is to assist the muftis and facilitate their job.

Searching in a dataset remains a fundamental problem for many applications. The general purpose of many similarity measures is to focus the search on as few elements as possible to find the answer. The current indexing techniques divides the target dataset into subsets. However, in large amounts of data, the volume of these regions explodes, which will affect search algorithms. The research tends to degenerate into a complete analysis of the data set. In this paper, we proposed a new indexing technique called GHB-tree. The first idea, is to limit the volume of the space. The goal is to eliminate some objects without the need to compute their relative distances to a query object. Peer-to-peer networks (P2P) are superimposed networks that connect independent computers (also known as nodes or peers). GHB-tree has been optimized for secondary memory in peer-to-peer networks. We proposed a parallel search algorithm on a set of real machine. We also discussed the effectiveness of construction and search algorithms, as well as the quality of the index.

In this paper, we present an approach for investigating information cascades in social and collaborative networks. The proposed approach seeks to improve methods limited to the detection of paths through which merely exact content-tokens are propagated. For this sake, we adopt to leverage web intelligence to the purpose of discovering paths that convey exact content-tokens cascades, as well as paths that convey concepts or topics related to these content-tokens. Indeed, we mine sequence of actors involved in cascades of keywords and topics extracted from their posts, using simple to use restful APIs available on the web. For the evaluation of the approach, we conduct experiments based on assimilating a scientific collaborative network to a social network. Our findings reveal the detection of missed information when using merely exact content propagation. Moreover, we noted that the vocabulary of actors is preserved mostly in short cascades, where topics become a better alternative in long cascades.

Understanding user’s intention is at the core of an effective images retrieval systems. It still a significant challenge for current systems, especially in situations where user’s needs are ambiguous. It is in this perspective that fits our study.In this paper, we address the challenge of grasping user’s intention in semantic based images retrieval. We propose an algorithm that performs a thorough analysis of the semantic concepts presented in user’s query. The proposed algorithm is based on an ontology and takes into account the combination of positive and negative examples. The positive examples are used to perform generalization and the negative examples are used to perform specialization which considerably decrease the two famous problems of image retrieval: noise and miss.Our algorithm processed in two steps: in the first step, we deal only with the positive examples where we will generalize the query from the explicit concepts to infer the others hidden concepts desired by the user. whereas the second step deal with the negative examples to refine results obtained in the first step. We created an image retrieval system based on the proposed algorithm. Experimental results show that our algorithm could well capture user’s intention and improve significantly precision and recall.

Statistical extractive summarization is one of the most exploited approach in automatic text summarization due to its generation speed, implementation easiness and multilingual property. We want to improve statistical sentence scoring by exploring a simple, yet powerful, property of graphs called bushy paths represented by the number of node’s neighbors. A graph of similarities is constructed in order to select candidate sentences. Statistical features such as sentence position, sentence length, term frequency and sentences similarities are used to get a primary score for each candidate sentence. The graph is used again to enhance the primary score by using bushy paths property. Also, we tried to exploit the graph in order to enhance summary’s coherence. We experimented our method using MultiLing’15 workshop’s corpora for multilingual single document summarization. Using graph properties can improve statistical scoring without loosing the multilingualism of the method.

This paper investigates feature selection method using two hybrid approaches based on artificial Bee colony ABC with Particle Swarm PSO algorithm (ABC-PSO) and ABC with genetic algorithm (ABC-GA). To achieve balance between exploration and exploitation a novel improvement is integrated in ABC algorithm. In this work, particle swarm PSO contribute in ABC during employed bees, and GA mutation operators are applied in Onlooker phase and Scout phase. It has been found that the proposed method hybrid ABC-GA method is competitive than exiting methods (GA, PSO, ABC) for finding minimal number of features and classifying WDBC, colon, hepatitis, DLBCL, lung cancer dataset. Experimental results are carried out on UCI data repository and show the effectiveness of mutation operators in term of accuracy and particle swarm for less size of features.

Light electric vehicle LEV autonomous present a major important problem for modern commercialized Electric Vehicle propulsion system. To improve the perfomance of LEV an efficiency fuzzy logic controller power management are proposed. The proton exchange mebran fuel hybrid system considered in this paper consists of fuel cells, lithium-ion batteries, and supercapacitors. The LEV is moving in the Algerian Saharan region, exactly in Bechar city. The aim objective of this work is to study the comportment of 2WDLEV based direct torque control supplied by differents sources of energy under diffrents speed varaiation. The performances of the proposed strategy controller give a satisfactory simulation results. The proposed control law increases the utility of LEV autonomous under several speed variations. Moreover, the future industrial’s vehicle must take into considerations the hybrid power management choice into design steps. ...

Verifying the authenticity of handwritten signatures is required in various current life domains, notably with official contracts, banking or financial transactions. Therefore, in this paper a novel histogram-based descriptor and an improved classification of the bio-inspired Artificial Immune Recognition System (AIRS) are proposed for handwritten signature verification. Precisely, the Histogram Of Templates (HOT) is introduced to characterize the most widespread orientations of local strokes in handwritten signatures, while the combination of AIRS and SVM is proposed to achieve the verification task. Usually, using the k Nearest Neighbor rule, a questioned signature is classified by computing dissimilarities with respect to all AIRS outputs. In this work, using these dissimilarities, a second round of training is achieved by the SVM classifier to further improve the discrimination power. In comparison with existing methods, the experiments on two widely-used datasets show the potential and the effectiveness of the proposed system.

In recent years, we are witnessing the spread of many and various modern real-time applications implemented on computer networks such as video conferencing, distance education, online games, and video streaming. These applications require the high quality of different network resources such as bandwidth, delay, jitter, and packet loss rate. In this paper, we propose an improved quantum chaotic animal migration optimization algorithm to solve the multicast routing problem (Multi-Constrained Least Cost MCLC). We used a quantum representation of the solutions that allow the use of the original AMO version without discretization, as well as improving AMO by introducing chaotic map to determine the random numbers. These two contributions improve the diversification and intensification of the algorithm. The simulation results show that our proposed algorithm has a good scalability and efficiency compared with other existing algorithms in the literature.

Because of the vital need for software systems to evolve and change over time in order to account for new requirements, software evolution at higher levels of modeling is considered as one of the main foundation within software engineering used to reduce complexity and ensure flexibility, usability and reliability. In similar studies for migration technique and software engineering, presenting a framework do not usually cover the specification of systems based on software architecture. In this paper, we specify a conceptual framework based on six explicit dimensions in respect to an architectural view-point as first class citizen. Indeed, sketching evolution relies upon identifying dimensions on which researchers try to answer while performing a new approach. The proposed model is based on answering What, Where, When, Who, Why and How questions. Analyzing these dimensions could provide a multiple choice to implement classification for architectural techniques. Further and using an example, these dimensions are quantified and then analyzed. This framework aims to provide a blueprint to guide researches to position architectural evolution approaches and maps them according a selected set of dimensions.

Nature-inspired algorithms attract many researchers worldwide for solving the hardest optimization problems. One of the well-known members of this extensive family is the evolutionary strategy ES algorithm. To date, many variants of this algorithm have emerged for solving continuous as well as combinatorial problems. One of the more promising variants, a self-adaptive evolutionary algorithm, has recently been proposed that enables a self-adaptation of its control parameters. In this paper, we discuss and evaluate popular common and self-adaptive evolutionary strategy (ES) algorithms. In particular, we present an empirical comparison between three self-adaptive ES variants and common ES methods. In order to assure a fair comparison, we test the methods by using a number of well-known unimodal and multimodal, separable and non-separable, benchmark optimization problems for different dimensions and population size. The results of this experiments study were promising and have encouraged us to invest more efforts into developing in this direction.

In this paper, three recent swarm intelligence algorithms (spider monkey optimization (SMO), social spider optimization (SSO) and teaching learning based optimization (TLBO)) are proposed to the optimization of microwave filter (H-plane three-cavity filter). The results of convergence and optimization use of these algorithms are compared with the results of the most popular swarm intelligences algorithm, namely particle swarm optimization (PSO) for different common parameters (population size and maximum number of iteration). The results showed validation of the proposed algorithms.

The innovations diffusion is a process of communication by which a new idea spreads among a population. A successful propagating leads researchers to seek the elements, which approve and consequently contribute to its spread, or otherwise, find the elements that prevent it. In fact, many efforts were made to models this social phenomenon; however, each one had its strengths and weaknesses. Therefore, this paper aims (1) to discuss and to analyze existing models showing their utility and limitation, (2) highlighting the detail of their applications, and (3) suggesting a taxonomy, which resumes the state of art. Then, to address the current social networks models, it seems necessary to begin by presenting the diffusion of innovation theory, then to detail models, which do not incorporate the social structure, more precisely the mathematical models.

In this research work, we propose a cooperative approach called simulated particle swarm optimization (SPSO) which is based on metaheuristics to find an approximate solution for the multiple sequence alignment (MSA) problem. The developed approach uses the particle swam optimization (PSO) algorithm to discover the search space globally and the simulated annealing (SA) technique to improve the population leader «gbest» quality in order to overcome local optimum problem. Simulation results on BaliBASE benchmarks have shown the potent of the proposed method to produce good quality alignments comparing to those given by other existing methods.

In this paper, we propose a semantic approach for automatic ontology learning from heterogeneous relational databases in order to facilitate their integration. The semantic enrichment of heterogeneous databases, which cover the same domain, is essential to integrate them. Our approach is based on Wordnet and Wup’s measure for measuring the semantic similarity between elements of these databases. It is described by a detailed process that can allow not only the generation of ontology but also its evolution as the evolution of its databases. We applied our approach in the alimentation risks field that is characterized by a large number of scientific databases. The developed prototype has been compared with similar tools of generation ontology from databases. The result confirms the quality of our prototype that returns the generic ontology from many relational databases.

Soft-biometrics prediction from handwriting analysis is gaining a wide interest in writer identification since it gives additional knowledge about the writer like its gender (man or woman), its handedness (left-handed or right-handed) and its age range. All research works developed in this context were focused on predicting a single soft-biometric trait. Nevertheless, it could be more interesting to develop a system that predicts several traits from a handwritten text. Presently, we investigate the feasibility of such multiple trait prediction. To reach this end, we propose two prediction schemes. The first combines individual prediction scores to aggregate a global prediction. The second scheme is based on a multi-class prediction. For both schemes, the prediction is based on SVM classifier associated with Gradient features. Experimental corpus is collected from IAM handwritten database. Conclusively, the second scheme proved to be more promising and evinced that the age characteristic is stable over time for a certain category of writers.

The enhancement of the machine condition monitoring process is a key issue for reliability improvement. In fact, in order to produce quickly, economically, with high quality while decreasing the risk of production break due to a machine stop, it is necessary to maintain the equipment in a good operational condition. This requirement can be satisfied by implementing appropriate maintenance strategies such as Condition Based Maintenance (CBM) and using updated condition monitoring technologies for faults detection and classification. In this context, a new method for machinery condition monitoring based on Mel-Frequency Cepstral Coefficients (MFCCs) and Support Vector Machine (SVM) is proposed to automatically detect the mechanical faults by maximized the generalization ability. Hence, the purpose is to design an automatic detection system for mechanical components defects based on supervised classification by trained to maximize the margin. The proposed approach consists in a sequence of binary classifications after extracting a set of relevant features such as temporal indicators and MFCC coefficients. The diagnosis accuracy assessment is carried out by conducting various experiments on acceleration signals collected from a rotating machinery under different operating conditions.

Clinical information systems store a large amount of data in medical databases. In the use of medical dataset for diagnosis, the patient’s information is selectively collected and interpreted based on previous knowledge for detecting the existence of disorders. Feature selection is important and necessary data pre-processing step in medical data classification process. In this work, we propose a wrapper method for feature subset selection based on a binary version of the Firefly Algorithm combined with the SVM classifier, which tries to reduce the initial size of medical data and to select a set of relevant features for enhance the classification accuracy of SVM. The proposed method is evaluated on some medical dataset and compared with some well-known classifiers. The computational experiments show that the proposed method with optimized SVM parameters provides competitive results and finds high quality solutions.

Virtual screening (VS) is an in-silico tool for drug discovery that aims to identify the candidate drugs through computational techniques by screening large libraries of small molecules. Various ligand and structure-based virtual screening approaches have been proposed in the last decades. Machine learning (ML) techniques have been widely applied in drug discovery and development process, predominantly in ligand based virtual screening approaches. Ensemble learning is a very common paradigm in ML field, where many models are trained on the same problem’s data, to combine in the end the results in one improved prediction. Applying VS to massive molecular libraries (Big Data) is computationally intensive; so the split of these data to chunks to parallelize and distribute the task became necessary. For many years, MapReduce has been successfully applied on clusters to solve the problems with very large datasets, but with some limitations. Apache Spark is an open source framework for Big Data processing, which overcomes the shortcomings of MapReduce. In this paper, we propose a new approach based on ensemble learning paradigm in Apache Spark to improve in terms of execution time and precision the large-scale virtual screening. We generate a new training dataset to evaluate our approach. The experimental results show a good predictive performance up to 92% precision with an acceptable execution time.

We present in this paper an easier way to manage activity traces and to compute human learning indicators activities in Technology Enhanced Learning TEL systems. We review our research work related to Trace based system TBS and we explain how we use TBS to develop new and generic model to represent the indicator life cycle from its creation to its reuse. This paper presents the underlying theory and how this theory is implemented to compute human learning indicators activities available for use with any other learning platform, provided the TBS can access the learning platform traces.

In this paper, a social-based collaborative filtering model named SBCF is proposed to make personalized recommendations of friends in a social networking context. The social information is formalized and combined with the collaborative filtering algorithm. Furthermore, in order to optimize the performance of the recommendation process, two classification techniques are used: an unsupervised technique applied initially to all users using the Incremental K-means algorithm and a supervised technique applied to newly added users using the K-Nearest Neighbors algorithm (K-NN). Based on the proposed approach, a prototype of a recommender system is developed and a set of experiments has been conducted using the Yelp database.

A whole classification process is the result of a discovery process that requires constant back and forth between theoretical description of the solution, software implementation, testing and refinement of the theoretical description in the light of the results of experimentation. This process is iterative. It should be, always, under the control of the user according to his subjectivity, his knowledge and the purpose of his analysis. In the last years, several platforms for digging data where classification is the main functionality have emerged. Some of these platforms allow a rapid prototyping and support a re-use of existing “computational modules” from existing “computational tool cases”. However, they lack flexibility and sound formal foundations. We propose, in this paper, a formal model with strong logical foundations, based on typed applicative systems. In this model, “computational modules” are considered as operators followed by their operands. A specific processing chain becomes a specific arrangement of a set of modules according to the needs of the user. The model ensures a firm compositionality of this arrangement.

In user memory based collaborative filtering algorithm, recommendation quality depends strongly on the neighbors selection which is a high computation complexity task in large scale datasets. A common approach to overpass this limitation consists of clustering users into groups of similar profiles and restrict neighbors computation to the cluster that includes the target user. K-means is a popular clustering algorithms used widely for recommendation but initial seeds selection is still a hard complex step. In this paper a new genetic algorithm encoding is proposed as an alternative of k-means clustering. The initialization issue in the classical k-means is targeted by proposing a new formulation of the problem, to reduce the search space complexity affect as well as improving clustering quality. We have evaluated our results using different quality measures. The employed metrics include rating prediction evaluation computed using mean absolute error. Additionally, we employed both of precision and recall measures using different parameters. The obtained results have been compared against baseline techniques which proved a significant enhancement.

Drug repositioning or repurposing refers to identifying new indications for existing drugs and clinical candidates. Predicting new drug-target interactions (DTIs) is of great challenge in drug repositioning. This tricky task depends on two aspects. The volume of data available on drugs and proteins is growing in an exponential manner. The known interacting drug-target pairs are very scarce. Besides, it is hard to select the negative samples because there are not experimentally verified negative drug-target interactions. Many computational methods have been proposed to address these problems. However, they suffer from the high rate of false positive predictions leading to biologically interpretable errors. To cope with these limitations, we propose in this paper an efficient computational method based on deep semi-supervised learning (DeepSS-DTIs) which is a combination of a stacked autoencoders and a supervised deep neural network. The objective of this approach is to predict potential drug targets and new drug indications by using a large scale chemogenomics data while improving the performance of DTIs prediction. Experimental results have shown that our approach outperforms state-of-the-art techniques. Indeed, the proposed method has been compared to five machine learning algorithms applied all on the same reference datasets of DrugBank. The overall accuracy performance is more than 98%. In addition, the DeepSS-DTIs has been able to predict new DTIs between approved drugs and targets. The highly ranked candidate DTIs obtained from DeepSS-DTIs are also verified in the DrugBank database and in literature.

The Multi-Robot Task Allocation problem (MRTA) is the situation where we have a set of tasks and robots; then we must decide the assignments between robots and tasks in order to optimize a certain measure (e.g. allocate the maximum number of tasks...). We present an effective solution to resolve this problem by implementing a two-stage methodology: first a global allocation that uses the Firefly Algorithm (FA), next a local allocation that uses the set theory properties (Power Set algorithm). Finally, results of the different simulations show that our solution is efficient in terms of the rate of allocated tasks and the calculated allocations are locally optimal.

Feature selection is a key issue in machine learning and data mining. A great deal of effort has been devoted to static feature selection. However, with the assumption that features occur over time, methods developed so far are difficult to use if not applicable. Therefore, there is a need to design new methods to deal with streaming feature selection (SFS). In this paper, we propose the use of dynamic optimization to handle the dynamic nature of SFS with the ultimate goal to improve the quality of the evolving subset of selected features. A hybrid model is developed to fish out relevant features set as unnecessary by an online feature selection process. Experimental results show the effectiveness of the proposed framework compared to some state of the art methods.

Optimization is known as the process of finding the best possible solution to a problem given a set of constraints. The problem becomes challenging when dealing with conflicting objectives, which leads to a multiplicity of solutions. Evolutionary algorithms, which use a population approach in their search procedures, are advised to suitably solve the problem. In this article, we present an approach for an evolutionary combinatorial multi-objective optimization of business process designs using a variation of NSGAII, baptized MA-NSGAII. The variants of NSGAII are numerous. In fact, the vast majority deals either with the crossover operator or with the crowding distance. We discuss an optimization Framework that uses (i) a proposal of effective Fitness function, (ii) 02 contradictory criteria to optimize and (iii) an original selection technique. We test the proposed Framework with a real life case of multi-objective optimization of business process designs. The obtained results clearly indicate that an effectual Fitness function combined with the appropriate selection operator affects undeniably quality and quantity of solutions.

Increasingly, the adoption of mobile sensors becomes imperative in the context of target tracking applications, especially for reliable data collection purpose. However, the design of a strategy that allows mobile sensors suitably to move in an autonomous, distributed and self-organized way is not evident to achieve by a deterministic polynomial algorithm. Solutions that are biologically inspired by the collective behaviour of individual social communities provide alternative tools and efficient algorithms that emerge from many interesting properties applicable to sensor technology. These solutions implement highly efficient systems that are structurally simple, powerful, highly distributed and fault-tolerant. Some biological societies, like colonies of the Escherichia coli bacteria, offer prospects to certain mobile sensors to acquire an artificial intelligence allowing them to move autonomously through the network. In this paper, we proposed a bio-inspired protocol named SMBFOA (Sink Mobility based on Bacterial Foraging Optimization Algorithm). The main idea of this protocol was inspired by the autonomous movement of the Escherichia coli bacterium. Based on the simulation results, we concluded that our proposed SMBFOA protocol increases the throughput data rate and prolongs the network lifetime duration for 30% and 5% respectively compared to Clustering Duty Cycle Mobility aware Protocol (CDCMP).

In this paper, Social-Spider optimization (SSO) algorithm is proposed for training artificial neural networks (ANN). Further, the trained networks are tested on two microwave filters modeling (Broad-band E-plane filters with improved stop-band and H-plane waveguide filters considering rounded corners). To validate the effectiveness of this proposed strategy, we compared the results of convergence and modeling obtained with the results obtained by NN used a population based algorithm namely Particle Swarm Optimization (PSO-NN). The results prove that the proposed SSO-NN method has given better results.

In the last decade, stochastic and meta-heuristic algorithms have been extensively used as intelligent strategies to resolve different combinatorial optimization problems. Honey Bee Mating Optimization is one of these most recent algorithms, which simulate the mating process of the queen of the hive. The scheduling algorithm is of paramount importance in a real-time system to ensure desired and predictable behavior of the system. Within computer science real-time systems are an important while often less known branch. Real-time systems are used in so many ways today that most of us use them more than PCs, yet we do not know or think about it when we use the devices in which they reside. Finding feasible schedules for tasks running in hard, real-time computing systems is generally NP-hard. In this work, we are interested in hybridizing this HBMO algorithm with other metaheuristics: Genetic Algorithms (GA), Greedy Random Adaptive Search Procedure (GRASP), Tabu Search (TS) and Simulated Annealing (SA) to resolve a real-time scheduling problem and obtain the optimal tasks schedule with respecting all temporal constraints. This is a complex problem which is currently the object of research and applications. In this scheduling problem, each task is characterized by temporal, preemptive and static periodicity constraints. The quality of the proposed procedure is tested on a set of instances and yields solutions which remain among the best.

Traditionally, process planning and scheduling functions are performed sequentially, where scheduling is implemented after process plans has been generated. Recent research works have shown that the integration of these two manufacturing system functions can significantly improve scheduling objectives. In this paper, we present a new hybrid method that integrates the two functions in order to minimize the makespan. This method is made up of a Shifting Bottleneck Heuristic as a starting solution, Tabu Search (TS) and the Kangaroo Algorithm metaheuristics as a global search. The performance of this newly hybrid method has been evaluated and compared with an integrated approach based on a Genetic Algorithm. Thereby, the characteristics and merits of the proposed method are highlighted.

In the last years, many hybrid metaheuristics and heuristics combine one or more algorithmic ideas from different metaheuristics or even other techniques. This paper addresses the hybridization of a primitive ant colony algorithm inspired from the Pachycondyla apicalis behavior to search prey with the Genetic Algorithm to find near optimal solutions to solve the Flow Shop Scheduling Problem with makespan minimization. The developed algorithm is applied on different flow shop examples with diverse number of jobs. A sensitivity analysis was performed to define a good parameter choice for both the hybrid metaheuristic and the classical Genetic Algorithm. Computational results are given and show that the developed metaheuristic yields to a good quality solutions.

To date, the multiproduct pipeline transportation mode has nationally and internationally considerably evolved thanks to his efficiently and effectively of transporting several products. In this paper, we focus our study on the scheduling of a multiproduct pipeline system that receives a number of petroleum products (fuels) from a single refinery source in order to be distributed to several storage and distribution centers (depots). Mixed Integer Linear Programming (MILP) continuous mathematical approach is presented to solve this problem. The sequence of injected products in the same pipeline should be carefully studied, in order to meet market demands and ensure storage autonomy of the marketable pure products in the fuels depots on the one hand and to minimize the number of interfaces; Birth zone of mixture between two products in contact and in sequential flow, which may hinder the continuous operation of the pipeline system, by the necessity of additional storage capacity for this last mixture, that is in no way marketable and requires special processing operations. This work is applied on a real case of a multiproduct pipeline that feeds the western and southwestern region of Algeria with fuels. The obtained results based on the MILP continuous approach give an optimal scheduling of the multiproduct transport system with a minimized number of interfaces.

This paper addresses the NP-hard problem of deploying wireless sensor networks on 3D terrains. On the contrary to previous works that place the sensors without any analysis of the terrain, we propose a two-phase solution based on terrain partitioning. The main idea is to estimate the number of sensors to be used and simplify the sensors deployment by partitioning the terrain according to topographic criteria. Simulation results based on real-world terrains confirm the efficiency of our solution in terms of coverage quality.

In this paper, we cope with malicious nodes dropping packets to disrupt the well-functioning of mobiles ad hoc networks tasks. We propose a new hybrid acknowledgment punishment scheme based on Dempster Shafer theory, called HAPS. The proposed scheme incorporates three interactive modules. The monitor module monitors the behaviour of one-hop nodes in the data forwarding process. The reputation module assesses the direct and the indirect reputation of nodes using Dempster Shafer theory, which is a mathematical method, that can aggregate multiple recommendations shared by independent sources, while some of these recommendations might be unreliable. Since recommendations exchange between nodes consumes resources, a novel recommendation algorithm has been incorporated to deal with false dissemination attack and to minimize the recommendation traffic. The exclusion module punishes nodes regarded as malicious. The simulation results show that HAPS improves the throughput and reduces the malicious dropping ratio in comparison to existing acknowledgment scheme.

VANET networks consist of several vehicles communicate with each other or with road fixed stations in order to offer a secure collaborative driving environment with the help of broadcasting emergency messages. These messages play a significant role in road safety by warning vehicles of any potential danger. The work reported in this paper is to devise an efficient broadcast protocol fulfilling alert message requirements in the VANETs context. Reacting in time to these urgent events may be important for crucial situations like road accidents.Performance evaluation of our proposal EBAM is carried out through simulations with combination of VanetMobiSim and Network Simulator 2 and obtained results, in terms of loss rate, delay and reception rate of urgent messages, outperform those given by the two known referred protocols.

The mainspring of a Wireless Sensor Network (WSN) is to capture data from an interesting deployment area and sending them out to the end user. However, one can ask to what extent the end user can have confidence in the use of these data? Especially when these collected data are employed in a crucial application that definitely excludes wrong data use. Thereby, WSN mission success is basically dependent on trustworthy of the data delivery process to the end user. To reach this goal, some obstacles, related to malicious node behavior or failure nodes, must be avoided or tolerated. Therefore, we propose, in this paper, a scheme to improve the dependability of the data sensed by sensor nodes in one hand and the reliable communication of these data to the sink in other hand. The proposal is based on a fault tolerant sensing process and the resilience to malware threat on the transmitted data from nodes to sink. The proposal was integrated to the well know Leach protocol and the performance evaluation, carried out on NS2 simulator, showed convincing results in terms of energy conserving, received data rate and node failure occurrence and attack detections.

Vehicular Ad Hoc NETworks (VANETs) has been developed to be, as a part of the Internet of Vehicles (IoVs), one of the most promoting technologies in the near future. Extensive researches are focusing on developing appropriate solutions in order to provide better Quality of Services (QoS) to these latter. This paper deals with network performance improvement based on clustering mechanism to reduce the number of communications caused by the high frequency of arrival/departure of vehicles while maintaining relevant clock synchronization solutions adequately operating. Thus, we attempt to adopt certain known clustering algorithms to our previous solution of time synchronization in VANETs namely OTRB (Offsets Table Robust Broadcasting) in order to evaluate the impact of a hierarchical communication on time synchronization process. An analytical study with a comparison on the influence of the clustering mechanism is given. The comparison takes place using simulation software NS2 (Network Simulator 2) and VanetMobiSim (VANET Mobility Simulator). The performance parameters include the average of arrival/departure of nodes and the number of isolated nodes. Simulation results reveal that an appropriate method should reduce the overhead of re-clustering and lead to an efficient network coverage with the best time synchronization rate.

Sensors play a vital role in Internet of Things (IoT) monitoring applications. However, the harsh nature of the environment influences negatively on the sensors reliability. They may occasionally generate inaccurate measurements when they are exposed to high level of humidity, temperature, etc. Forwarding incorrect data to the base station may cause disastrous decision-making in critical applications. To address this problem, we proposed a new fault-tolerant mechanism based on Bayesian Networks that carefully solve such a problem. The mechanism is called Reliability of Captured Data (RCD); it enables sensors to both detect and recover faulty data. To show the performance of our algorithm, we compared this latter with another recent algorithm called Fault Detection Scheme (FDS) using Network Simulator 2 (NS2). The results showed that our strategy gives remarkable enhancements in terms of fault detection accuracy and fault alarm rate.

In smarts cities environments, a recommender system (RS) has for goal to recommend relevant services to the user who is sometimes mobile. Thus, to be able to provide accurate personalized recommendations, the RS should be aware to the user’s context (preferences, location, activities, environment, ...), thereby, it should be Context-Aware Recommender System (CARS, for short). Therefore, the context modeling becomes crucial for developing CARSs. Although there is a lack of context models in the RS literature, several ones have been proposed in pervasive computing field. Nevertheless, most of them are dedicated for closed spaces and should be reviewed to be more suitable for open intelligent environments such as smart cities. This paper aims to propose an extensible ontology-based context model for representing contextual information within a smart city. The proposed context model would subsequently allow to design and develop CARSs.

The Web of Things (WoT), facilitates the interconnection of different types of real-world objects, integrating them into the virtual world and ensuring their interoperability through Web services. However, it remains a challenge to automate the tasks connected objects need to deal with. In this paper, we focus on the development of smart web services that automate service tasks, autonomously adapt to context changes in the object’s environment, and to users’ preferences. In this article, we propose a software framework for smart services that relies on a reactive and proactive approach to deal with context and its temporal aspects. Smart web services developed according to these principles can react to current situations and proactively anticipate an unforeseen situation in order to take the right decision.

Service composition is seen as the key issue to create innovative, efficient, flexible and dynamic applications on the Internet of things (IoT). Accordingly, we propose, in this paper, an approach for IoT service composition based on multi-agent system where several agents are engaged to satisfy the user request. This approach is designed using SysML and implemented using Netlogo platform. The use-cases scenarios and extensive tests show clearly the interest, the feasibility and the suitability of the multi-agent system for service composition.

Dealing with crisis situations involves significant collaborative decision-making to recover lives and preserve properties. In order to reach timely and appropriate decisions, crisis response organizations need to rapidly obtain an accurate situation awareness of the crisis context. A process by which they need to gather, access, and exchange near real-time information about the events circumstances throughout the entire crisis life cycle. Such activities imply a huge participation of collaborating organizations and goes even beyond their internal borders to reach the entire crisis-stricken community. Recent crisis and disaster situations have demonstrated the crucial role citizens can and must play in responding to such events. The growing development of advanced technologies has open the door for large public to be a key factor in making decisions and conducting guided response actions. We propose in this paper a comprehensive approach that integrates crisis crowdsourcing tasks and techniques to crisis decision-making activities. To demonstrate its relevance, we present a study of crisis scenario based on the context of Desert Locust Plague Survey and Control in the Algerian National Institute of Plant Protection (INPV).

Face recognition has been very popular in recent years, for its advantages such as acceptance by the wide public and the price of cameras, which became more accessible. The majority of the current facial biometric systems use the visible spectrum, which suffers from some limitations, such as sensitivity to light changing, pose and facial expressions. The infrared spectrum is more relevant to facial biometric, for its advantages such as robustness to illumination change. In this paper, we propose two multispectral face recognition approaches that use both the visible and infrared spectra. We tested the new approaches with Uniform Local Binary Pattern (uLBP) as a local descriptor and Zernike Moments as a global descriptor on IRIS Thermal/Visible and CSIST Lab 2 databases. The experimental results clearly demonstrate the effectiveness of our multispectral face recognition system compared to a system that uses a single spectrum.

Image segmentation is the process of partitioning the image into regions of interest in order to provide a meaningful representation of information. Nowadays, segmentation has become a necessity in many practical medical imaging methods as locating tumors and diseases. Hidden Markov Random Field model is one of several techniques used in image segmentation. It provides an elegant way to model the segmentation process. This modeling leads to the minimization of an objective function. Conjugate Gradient algorithm (CG) is one of the best known optimization techniques. This paper proposes the use of the nonlinear Conjugate Gradient algorithm (CG) for image segmentation, in combination with the Hidden Markov Random Field modelization. Since derivatives are not available for this expression, finite differences are used in the CG algorithm to approximate the first derivative. The approach is evaluated using a number of publicly available images, where ground truth is known. The Dice Coefficient is used as an objective criterion to measure the quality of segmentation. The results show that the proposed CG approach compares favorably with other variants of Hidden Markov Random Field segmentation algorithms.

We present in this paper, a facial expressions recognition system to command a mobile robot (Pionner-3DX). The proposed system mainly consists of two modules: facial expression recognition and robot command. The first module aims to recognize the facial expressions like happiness, sadness, surprise, anger, fear, disgust and neutral using Gradient Vector Flow (GVF) snake to find ROI (Region Of Interest like: mouth, eyes, eyebrow) segmentation from FEEDTUM database (video file). While the second module, analyses the segmented ROI to recognize with Euclidian distance calculation (compatible with the MPEG-4 description of the six universal emotions) and Time Delay Neural Network classifier. Finally, the recognized facial expressions were used as control commands for the mobile robot displacement (forward; backward; turn left; turn right) in ROS (Robot Operating System).

In this paper, we give the design and implementation of a system for person re-identification in a camera network, based on the appearance. This system seeks to construct an online database that contains the history of every person that enters the field of view of the cameras. This system is qualified to associate an identifier to each detected person, which keeps this identifier in the same camera and in other cameras even if he or she disappears and then appears again. Our system comprises a moving objects detection step that is implemented using the Mixture of Gaussians method and a proposed difference method, to improve the detection results. It also comprises a tracking step that is implemented using the sum of squared differences algorithm. The re-identification stage is realized using two steps: the intersection of tracking and detection for the temporal association, the histogram for comparison. The global system was tested on a real data set collected by three cameras. The experimental results show that our approach gives very satisfactory results.

Because they are exploited in many fields such as medicine, agriculture, chemistry and others, plants are of fundamental importance to life on earth. Before it can be used, a plant need to firstly be identified and categorized. However, a manual identification task requires time, and it is not an easy task to do. This is because some plants look visually similar to the human eye, whereas some others may be unknown to it. Therefore, there has been an increasing interest in developing a system that automatically fulfils such tasks fast and accurate. In this paper, we propose an automatic plant classification system based on a parallel combination technique of multiple classifiers. We have considered using three widely known classifiers namely Naïve Bayes (NB), K-Nearest Neighbour (KNN) and Support Vector Machine (SVM). Our system has been evaluated using the well-known Flavia dataset. It has shown a better performance than those obtained using only one classifier.

Recommender systems are very useful to help access to relevant information on the web and to customize search. Content based filtering (CBF) is an alternative among others used to design recommender systems by exploiting items’ contents. Basically, they recommend items based on a comparison between the content of items and user profile. Usually, the content of an item is represented as a set of descriptors or terms; typically the words that occur in text documents. The user profile is represented by the same terms and built up by analysing the content of items he used before. However current CBF recommender systems are mostly devoted to deal with textual resources and cannot be used in their current form to handle the variety of data published on the web especially unstructured data. Another challenge for the existing CBF methods is the issue of new user for whom the system cannot draw any inference due to the lack of information about the user. This paper describes an approach to CBF that aims to deal with these problems on which CBF systems perform poorly. The basic feature of the proposed approach is to incorporate linked data cloud into the information filtering process using a semantic space vector model, and FOAF vocabulary, which is used to define a new distance measure between users, based on their FOAF profiles. We report on some experiments and very promising results of the proposed approach.

The uncertain QoS management is gaining a lot of interest in the service oriented computing area. In this work, we propose a framework that allows to select the TopK compositions of services that best meet the user’s requirements. This framework not only handles the user’s global constraints but it also takes into account the fluctuating nature of the QoS informations. More specifically we present two algorithms that ensure the aforementioned purposes. The first one ranks the services of each abstract class according to the probabilistic dominance heuristic. The second one explores the compositions search space by leveraging the backtracking search. The experimental evaluation shows that the proposed heuristic is more effective than the ranking based on average QoS.

Nowadays, current trends of universal quantification-based queries are been oriented towards flexible ones (tolerant queries and-or those involving preferences). In this paper, we are interested in universal quantification-like queries dealing with both positive or negative preferences (requirements or prohibitions), considered separately or simultaneously. We have emphasised the improvement of the proposed operator, by designing new variants of the classical Hash-Division algorithm, presented in [1], for dealing with our context. The parallel implementation is also presented, and the issue of answers ranking is dealt with. Computational experiments are carried out in both sequential and parallel versions. They shows the relevance of our approach and demonstrate that the new operator outperforms the conventional one with respect to performance (the gain exceeds a ratio of 40).

Efficient reasoning in intrusion detection needs to manipulate different information provided by several analyzers in order to build a reliable overview of the underlying monitored system trough a central security information and event management system (SIEM). SIEM provides many functions to take benefit of collected data, such as Normalization, Aggregation, Alerting, Archiving, Forensic analysis, Dashboards, etc. The most relevant function is Correlation, when we can get a precise and quick picture about threats and attacks in real time. Since information provided by SIEM is in general structured and can be given in XML, we propose in this paper to use an ontological representation based on Description Logics (DLs) which is a powerful tool for knowledge representation and reasoning. Indeed, Ontology provides a comprehensive environment to represent any kind of information in intrusion detection. Moreover, basing on DLs and rules, Ontology is able to ensure a decidable reasoning. Basing on the proposed ontology, an alert correlation prototype is implemented and two attack scenarios are carried out to show the usefulness of the semantic approach.

With the increase of volume, velocity, and variety of big data, the traditional collaborative filtering recommendation algorithm, which recommends the items based on the ratings from those like-minded users, becomes more and more inefficient. In this paper, two varieties of algorithms for collaborative filtering recommendation system are proposed. The first one uses the improved k-means clustering technique while the second one uses the improved k-means clustering technique coupled with Principal Component Analysis as a dimensionality reduction method to enhance the recommendation accuracy for big data. The experimental results show that the proposed algorithms have better recommendation performance than the traditional collaborative filtering recommendation algorithm.