IoT Fuzzy Logic Aquaponics Monitoring and Control Hardware Real-Time System
Adnan Shaout and Spencer G. ScottThe University of Michigan - Dearborn The Electrical and Computer Engineering Department Dearborn, MIAbstract—this paper presents a design for monitoring and controlling a fish tank and growing bed in an aquaponics ecosystem. Aquaponics is a growing field in which fish and plants are grown together and mutual benefit each other. Fuzzy logic is used to evaluate the inputs and automatically provide the proper output. The system will monitor water temperature, pH, air temperature, and luminance. The system will control a light, heater, and alarm. The Arduino Uno R3 board was selected to be the hardware interface for inputs/outputs. Selecting the Arduino was based on Matlab having a support toolbox to interface with the Arduino (ATMEGA8U2-MU) microcontroller. Updating of the input value and triggering twitter alerts was all done through using the free Thingspeak server tool that connects nicely with Matlab via a toolbox.
Keywords: Aquaculture, Aquaponics, Ecosystems, Fuzzy logic, MATLAB, Real-time systems, IoT, Arduino.
Real Time Hardware System for Synchronizing IoT Device State using the MQTT Protocol
Adnan Shaout and Brennan CrispinThe Electrical and Computer Engineering Department The University of Michigan -Dearborn
Abstract—with the growing number of IoT devices, there is an increasing need for the ability to synchronize state and data across multiple IoT devices. As many IoT devices will be operating in constrained environments with unstable network connections, understanding the tradeoffs of various protocols is critical. The primary objective of this paper is to design and implement an embedded system to connect to a M2M broker, use the cloud server to communicate with other embedded systems, and be configurable from a cloud-based web service. In this paper we explore previous research on machine to machine (M2M) protocols such as AMQP and MQTT, and demonstrate an MQTT based system for synchronizing IoT device state across multiple client nodes. The main goal of the system is for state changes to be registered and distributed throughout the system in under 1 second; and initial registration of a new node should occur in under 30 seconds.
Keywords: IoT, Machine to Machine, Embedded, MQTT, Cloud Computing, Synchronizing.
Applying an Algorithm on GSM Model Using PSK Constellations
1Jilali ANTARI, 2Mustapha BOUKRI and 1Abdessamad MALAOUI1Polydisciplinary Faculty Taroudan, Morroco 2LMTI,Faculty of Sciences Agadir, MorrocoAbstract—In this paper, we apply an algorithm based on fourth order cumulants (ALG-CUM4) on GSM (Global System Mobile) model using PSK (Phase-Shift Keying) constellations. The goal of the application is the equalization of communication channel in GSM. The simulation results and the comparison with another algorithm, also based on fourth order cumulants using ISI (Inter-Symbol Interference) criteria and PSK constellations, demonstrates the performances of the algorithm (ALG-CUM4) in different SNR (signal-to-noise ratio).
Keywords: Simulation; GSM model; Equalization; Fourth Order Cumulants.
Enhanced Multi-keyed Risk Adaptive Hybrid RFID Access Control System
Malek Al-Zewairi, Salam Hamdan and Mustafa Al FayoumiPrincess Sumaya University for Technology, JordanAbstract—the traditional access control system has been proven inefficient in dealing with modern systems where the access decision might be influenced by several factors. Several works have to be done on next-generation access control systems where risk plays a crucial role in the access control making the decision, which has led to propose several risk adaptive/aware access control models. In this paper, an enhanced multi-keyed model for generating the symmetric encryption key dynamically on the fly is proposed. The experimental results show that the proposed model has improved the overall security while preserving the same architecture of the previous model.
Keywords: RFID; Access Control; Risk; Fuzzy Logic.
A quality-aware context information selection based fuzzy logic in IoT environment
1Farida Retima, 2Saber Benharzallah , 1Laid Kahloul and 1Okba Kazar1University Mohamed Khider Biskra, Algeria 2Batna 2 University, Algeria
Abstract— In the last decade, several works proposed their own approaches about the context management for the internet of things (IoT). An important issue in such systems is faced by context data distribution with a sufficient level of quality i.e. quality of context (QoC). In this paper, a fuzzy logic-based framework is proposed which handles QoC evaluating within distributed context manager and context-aware applications. This article presents also a solutions provided by the MDE (Model Driven Engineering) for modeling the captured context information from different context source.
Keywords: IoT; context manager; MDE; fuzzy logic; QoC; context source; application.
A comparative analysis of IoT service composition approaches
Idir AOUDIA, Saber BENHARZALLAH, Laid KAHLOUL and Okba KAZARLINFI Laboratory, Biskra University 07000 Biskra, AlgeriaAbstract—The Internet of things is the integration of information space and physical space, becoming more and more popular in several places. A number of approaches to IoT's service composition have been proposed. In this paper, we will present a review of existing approaches for IoT's service composition; it describes and compares them among each other with respect to some key requirements. This paper also presents some basic concepts to introduce IoT challenges with a comparison between traditional Web service composition and IoT service composition. This paper represents a support for researchers to focus on their efforts and to deliver lasting solutions in this field.
Keywords: Internet of Things; service composition; adaptability; context.
Service composition based on the social relations in the Internet of things
Marwa Meissa, Saber BENHARZALLAH and Laid KAHLOULIntelligent Computer Science Laboratory/ Biskra University, AlgeriaAbstract: The Internet of things (IoT) is a collection of things internet worked via a network that enables to exchange the data. Furthermore, the object in IoT can be offer his functionalities as a physical or software services. The current trend of Web of things faces many challenges especially in the issue of the IoT service composition. However Previous published studies are limited to IOT service composition, there have been no controlled studies which analyze the recent works and very few studies have investigated the exploitation of the social relations; leveraging of the social relationships in composition phases can lead to address many challenges such as the navigability, the availability, and the scalability. Besides, it can establish the trustworthiness and enhance the effectiveness and the efficiency. Hence, in this paper, we aim to highlight on the IoT service composition challenges by analyzing previous studies. Another objective of this paper is to attempt to show the use of the social relation to enable the imposing the social dimension in the Web service composition.
Keywords: Service Composition, social relations, SIOT, Internet of Things service.
Formal Verification of Collision Free Mobility Adaptive Protocol for Wireless Sensor Networks
Manel Houimli and Laid KahloulLINFI Laboratory, Computer Science Department, Biskra University, AlgeriaAbstract—There has been a rapid growth in the need to support mobile nodes in Wireless Sensor Networks (WSNs). This may lead to several uncertainties and stochastic events such as messages loss, collisions on medium and even loss of nodes. Therefore, it is essential to ensure the efficiency of mobility based wireless sensor protocols. A lot of protocols have been proposed to consider mobility in the MAC layer as well as in the network layer. We opt to study formally one of these protocols using Statistical Model Checking (SMC). The chosen protocol is CFMA (Collision Free Mobility Adaptive) MAC protocol, for WSNs. This protocol performs well in both static and mobile scenarios. In this paper, we make a formal study for its principal algorithms by modeling, first, its work-flow using probabilistic timed automata. This stochastic formalism leads to consider several random events with more realism. Furthermore, a performance analysis is performed by verifying the qualitative and quantitative properties of this protocol using UPPAAL SMC tool.
Keywords: WSN; CFMA MAC protocol; Statistical Model Checking; UPPAAL SMC.
Reconfigurable Manufaturing Systems (RMSs): A literrature review on the use of genetic algorithms for RMSs
Fatima Zahra Torki, Laid Kahloul, 1Saber Benharzallah and Leila BelaicheLINFI Laboratory, Computer Science Department, Biskra University, Algeria1LINFI Laboratory, Computer Science Department, Batna 2 University, AlgeriaAbstract: in the 21st century, Manufacturing Systems (MSs) will face serious challenges. To stay competitive, to survive in the current context and to fulfill the unpredictable and frequent market changes driven by global competition represent some of these challenges. This paper deals with the new concept of manufacturing systems which address these challenges known as Reconfigurable Manufacturing Systems (RMS). RMSs launched firstly in 1999 at University of Michigan, have been widely studied in many scientific publications with or without industrial applications. The objective of the current paper is to make an overview highlighting the core characteristics of RMSs, presenting a comparison between different manufacturing system paradigms and finally illustrating the use of genetic algorithms in RMSs optimization.
Keywords: reconfigurable manufacturing system; reconfigurability; genetic algorithm; RMT; DMS; FMS.
Towards Building a Frame-Based Ontology for the Arabic Language
Mariam Biltawi, Arafat Awajan and Sara TedmoriPrincess Sumaya University for Technology, JordanAbstract— This paper proposes a framework for building a frame-based ontology for the Arabic language. The proposed framework consists of two main phases. The first phase involves manual construction of a seed frame-based ontology. This is followed by the second phase in which the seed frame-based ontology is enriched with new lexical fields (only if they do exist), and/or enriching it with binary relations between existing or new lexical fields. The binary relations considered are synonyms/antonyms, hyponyms/hypernyms, and holonyms/meronyms. In addition, the paper presents a comprehensive introduction of lexical semantics providing examples from the Arabic language, and then surveys works of researchers aiming to build ontologies for the Arabic language.
Keywords—Arabic ontology; lexical semantics; natural language processing.
Distinguishing Nominal and Verbal Arabic Sentences: A Machine Learning Approach
Duaa Abdelrazaq, Saleh Abu-Soud and Arafat AwajanPrincess Sumaya University for Technology JordanAbstract The complexity of Arabic language takes origin from the richness in morphology, differences and difficulties of its structures than other languages.Thus, it is important to learn about the specialty and the structure of this language to deal with its complexity. This paper presents a new inductive learning system that distinguishes the nominal and verbal sentences in Modern Standard Arabic (MSA).The use of inductive learning in association with natural language processing is a young and an interdisciplinary collaboration field, specifically in Arabic Language.The resultsobtained out of this research are ambitious and prove that implementing inductive learning in Arabic complex structure will gain a promisingcontribution in the field of Arabic natural language processing (ANLP).
Keywords: Arabic LanguageProcessing, Natural Language Processing, Inductive Learning, ILA.
Semantic Elements Extraction based on Syntactic Structure of Arabic Sentences
Raghda Hraiz, Mariam Khader, Arafat Awajan, Akram AlkousComputer Science Princess Sumaya University for TechnologyAmman, Jordan
Abstract— Huge amount of data is uploaded to the Internet daily, most of this data are a user-generated data. User-generated data impose challenges in processing by machines because of its unstructured nature. This paper proposes a new approach for extracting semantic relations from Arabic sentences. The proposed approach analyzes the syntactic structure of Arabic sentences and recognizes its semantic elements. The proposed approach consists of three stages: syntactic analysis, segmentation and semantic relation extraction phase. The extracted elements from the text are organized in XML format. The proposed approach has been evaluated using a dataset of 15 sentences from AL-Jazeera website. The results have shown that the proposed method is reliable and promising in extracting semantic elements.
Keywords—Semantic Network; Semantic Elements; Information Extraction; Natural Language Processing; knowledge Representation.
ATwo Stage Intrusion Detection Intelligent System
Nevrus Kaja, Adnan Shaout and Di MaThe University of Michigan – Dearborn, United StatesAbstract—Security is becoming an inherited and amplified problem that keeps growing as new and emerging technologies hit the market. The problem of detecting and preventing malicious attacks is just one of the puzzle pieces that many enterprises prioritize in their operations. Challenges brought by security breachesor malicious attacks can put a high stress, almost un-survivable, on social and economic factors. In this paper, the usage of a two stage intelligent system in intrusion detection systems (IDS) will be introduced. Anomaly based IDS' are a feasible solution for many security problems, but the high rate of false positives makes them difficult to implement. The objective of the first stage in this proposed IDSis to simply detect whether there is an attack present through unsupervised learning. The second stage classifiesthese attacks in a supervised learning methodology, along with addressing and eliminating the number of false positives. The simulation of this approach results is an IDS able to detect and classify attacks at a 99.97% accuracy and lowers the false positives rate to 0%.
Keywords—Intrusion detection systems (IDS), security, machine learnings, IPS, Knowledge discovery in databases (KDD).
RNN-LSTM based Beta-elliptic model for Online handwriting script identification
RAMZI ZOUARI, HOUCINE BOUBAKER and 1MONJI KHERALLAHNATIONAL SCHOOL OF ENGINEERS OF SFAX, TUNISIA1FACULTY OF SCIENCES OF SFAX, TUNISIAAbstract—Recurrent Neural Network has achieved the state-of-the-art performance in a wide range of applications dealing with sequential input data. In this context, the proposed system aims to classify the online handwriting scripts based on their labelled pseudo-words. To avoid the vanishing gradient problem, we have used a variant of recurrent neural network with Long Short-Term Memory. The representation of the sequential data is done through the beta-elliptic model. It allows extracting the kinematics and geometrics profiles of the different strokes constituting a script over the time. This system was assessed with a large vocabulary containing scripts from ADAB, UNIPEN and PENDIGIT databases. The experiments results show the effectiveness of the proposed system which achieved a recognition rate of 100% with only one recurrent layer and using the dropout technique.
Keywords—online; pseudo; stroke; velocity; beta-elliptic; recurrent; LSTM; dropout.
Conditional Arabic Light Stemmer: CondLight
Yaser A. M. Al-Lahham, Khawlah Matarneh and Mohammad A. HassanComputer Science Department , Zarqa UniversityZarqa – JordanAbstract Arabic language has a complex morphological structure, which makes it hard to select index terms for an IR system. The complexity of the Arabic morphology caused by multimode terms, using diacritics, letters have different forms according to its location in the word and affixes can be added at all locations in a word. Many methods were proposed to overcome this problem; such as root extraction and light stemming. Light stemming show better retrieval efficiency, Light10 is the best stemmer among a series of light stemmers, it simply removes suffixes and prefixes if it is listed in a predefined table. Light10 has no restrictions on the affixes, so it is possible to have two different terms having the same token while they have different meanings. This paper proposes adding extra prefixes and suffixes to the table, and imposes some conditions on removing these affixes. The implementation and testing of the proposed method show better precision than the Light10 stemmer.
Keywords: Arabic IR, Light stemming, term selection, Arabic document indexing.
A proposed framework for Botnet Spam-email Filtering using Neucube
1Ammar ALmomani, 2Mohammad Alauthman, 3Omar Almomani and 4Firas Albalas1IT-department, Al-Huson University College, Al-Balqa Applied University, Jordan
2Department of Computer Science, Faculty of information technology, Zarqa University, Jordan
3 Faculty of Information Technology The World Islamic Sciences & Education University, Jordan
4Jordan University of Science And Technology, Jordan
Abstract: The voluminous amount of SPAM emails that originated from different botnets worldwide has a direct effect on the limited capacity of mailboxes, security of personal mail, space-loss from the communication as well as the time that is required to identify and address the spam emails. These issues are crucial to the Internet and Networks OS that need a serious attention from the research community. As a result of such sophistication, attackers employ Botnet infrastructure for distributing spam email messages for malicious purposes. The proposed research will address the source of the above issues through identifying the SPAM emails before causing any further disturbance to the networks normal operations, due to the complexity of spamming techniques which are evolving from traditional spamming technologies (direct spamming) to more sophistication techniques (zero-day spam). Therefore, the proposed Online Botnet Spam E-mail Filtering Framework (BSEFF) uses the principles of a new spiking neural network architecture called Neucube algorithms [1, 2], we will focus on adaptive Dynamic Evolving Spiking Neural Network (deSNN) algorithm. The proposed algorithm is designed to handle large and fast spatio/spectro temporal data through applying the spiking neural networks (SNN) as the core processing module. To the best of our knowledge, the proposed framework can be considered as the first Spam detection approach that utilizes Neucube algorithm. BSEFF will inherit the feature of Neucube algorithm that adopts hybrid (supervised/unsupervised) learning approach, and use this algorithm in an online framework with long life learning inorder to classify input while the framework is learned.
Keywords: Botnet Spam-mail, NeuCubealgorithms, spiking neural networks, evolving spatio-temporal data machines (eSTDM).
Management of Construction and Maintenance Projects in Tunisian Public Sector : Brief Comparison Tools and Suggestions For Future Frameworks
Sirine Touati, 1Safa Bhar Layeb and Jouhaina Siala Chaouachi University Of Carthage, Tunisia1University of Tunis El Manar, TunisiaAbstract—Using Project Management (PM) techniques and tools has grown in developing countries and it has become an essential need especially in public sector. Tunisia, one of the growing economies in North Africa and the Middle East region since 1990s, implemented complex projects, which have a major impact on the development of the country especially in construction field. Using software tools seems useful in all the phases of project management cycle. This paper presents a brief comparison of popular PM tools used in different sectors and the ability of these tools to be adapted in Tunisian public sector for construction and maintenance projects. Suggestions for future frameworks are proposed.
Keywords—Project Management; Construction and Maintenance Projects; Public Sector; Project Management Software; Tunisia.
Impact of enterprise resource planning systems On Scientific Research System in Public University
Mohamed EL MOHADAB, Belaid BOUIKHALENE and Said SAFI Sultan Moulay Slimane University, MoroccoAbstract—scientific researches represent a very important axis for the university, because it ensures its innovation and its productivity and develop the competencies of these researcher and research laboratory. But the management and automation of this sector represents a great challenge for universities either for managers, directors, or the researchers from which comes the need of find relevant and effective solution. To manage this sector, we have to study several computer solutions which ensure good management of the information system but trying to use open source and scalable solutions to ensure adaptation to the new change that can be brought to the system.
Keywords— Enterprise Resource Planning; ERP systems; ERP implementation; Scientific Research; Public University; case study approach; Odoo.
Aware-Routing Protocol using Best First Search Algorithm In Wireless Sensor Networks
Ghassan SamaraInternet Technology Department Zarqa University, JordanAbstract Wireless Sensor Networks (WSNs), sometimes called Wireless Sensor-Actuator Networks (WSANs) are increasingly important because of their practical use in different aspects of our life; this led to proliferation wireless sensor networks at everywhere. Despite the rapid spreading of the WSNs, but energy issue is one of the most significant problems and challenges in this field. Recent developments and widespread in wireless sensor network have led to many routing protocols that try to mitigate this issue. WSNs consist of low-power sensor nodes and a few base stations (sink nodes), all these devices have to be adaptive and efficient in data transmission. This paper proposes an aware-routing protocol based on Best First Search Algorithm (AR-BFS), this protocol aims to focus on the energy efficiency, by reducing the powerconsumption of each sensor node, increase network lifetime, and ensure system reliability.
Keywords Best First Search (BFS), Heuristic Function, Wireless Sensor Networks (WSNs), Wireless Sensor Actuator Network (WSANs).
Integrate Database Design Techniques with Agile Applications
Alaa M. El-Halees and Emad O. KehailFaculty of Information Technology Islamic University of Gaza Gaza, PalestineAbstract— One of the major business needs nowadays is the ability to respond to business requirements and changes quickly. Therefore, the software development process has been moved to be more agile by using what has been agreed to call Agile Software Development Process. However, the business still needs to store data, and Database Management Systems (DBMS) are still the de facto for the business software. DBMS is relying completely on database design process that follows traditional up-front design process which is sequential by nature. This research developed a model that integrates the database design techniques with Scrum Agile practices. The new model did not sacrifice the features of the database design techniques, yet the model help to make the database design process more agile by distributing the database design process among the Scrum development process. We evolve our new model by using Focal Point approach and then adding an Abstraction Layer at the database level. We found that the new model helped to reduce the impact of the changes implemented at the database level and to achieve the goal with a percentage around 64% of the time needed to achieve the same goal using the traditional upfront design. This is in addition to the flexibility of the new system when it comes to adapt new changes since the results showed that the new model is around 80% more flexible than using upfront design approach.
Keywords— Agile Software Development, Database, SCRUM.
Support Vector Machine based Feature Selection Method for Text Classification
Thabit Sabbah, Mosab Ayyash and Mahmood AshrafFaculty of Technology and Applied Sciences, Al-Quds Open University, Ramallah, Palestine Department of Computer Science, Federal Urdu University of Arts, Science & Technology, Islamabad, Pakistan
Abstract—Automatic text classification is one of the most effective tools used to sort out the increasing amount of textual content available online. High dimensionality remains one of the major obstacles observed in the text classification field in spite of the fact that there have been statistical methods available to face this issue. Still, none of them has proved to be effective enough to solve this problem. This paper proposes a feature ranking and selection method based on the Support Vector Machine (SVM) learning algorithm, also known as (SVM-FRM). This method assumes that weights given by the SVM learning algorithm to different features in feature space indicate the significance of these features. As such, the feature selection process can be established based on the referred to weights. The researchers tested the SVM proposed method using three text classification public datasets. Then, they compared the results to those of other statistical feature selection methods currently used for this purpose. In the light of this comparison, applying the proposed SVM-FRM method for text classification has proved to have a superior F-measure and accuracy performances than the rest of other methods applied for this purpose, when tested on balanced datasets, in spite of its size and the high competing performances on an unbalanced dataset.
Keywords— Feature ranking; text classification; feature selection; SVM; dimensionality reduction.
A General Structure for the Approaches Used in the Human Action and Activity Recognition from Video
Nawaf Alsrehin1 ,Bilal Abul-Huda21Computer Information Systems Department, Yarmouk University, Jordan2Management Information Systems Department, , Yarmouk University, JordanAbstract: Recognizing human actions and activities from videos has become an important topic in computer vision, machine learning, and pattern recognition applications. Some of these applications include automatic video analysis, human behavior recognition, human machine interaction, robotics, and security aspects in real-time video surveillance. This paper provides a general review of most recent advances and approaches in human action and activity recognition during the past several years. It also presents a categorization of human action and activity recognition approaches and methods with their advantages and limitations. In particular, it divides the recognition process based on (a) the method used to extract the features from the input image/video, (b) learning classifier techniques. Moreover, it presents an overview of the existing and publically available human action and activity recognition datasets. This paper also examines the requirements for an ideal human recognition system and presents some directions for future research and report some exposed problems on human action and activity recognition.
Simplified Digital Logic Circuits Using Binary Decision Tree
Hamed A. Fawareh and Abla Suliman Hussiensoftware Engineering Department, ZarqaUniversity, JordanAbstract—Ordered Binary decision tree (OBDT) is a graphical representation which looks like a tree with root and branches; it played a key role in digital circuits verification and manipulation which leads to a compact circuit after elimination many redundant branches. In this paper a Boolean equation was created for the proposed digital circuit, and with the aid of Boolean Algebra fundamentals, the proposed Boolean equation was simplified and reduced in size, afterword a truth table was constructed from which a Binary Decision Tree was sketched, then followed with a procedure of reducing and eliminating redundancy in the branches, a final schematic BBD will be constructed, two case studies were presented for verification of this technique.
Index Terms—binary decision tree (BDT), digital logic circuit and systems, binary decision diagram (BDD, Binary Algebra.
Recognizing handwritten Arabic words using optimized character shape models and new features
1Anis Mezghani and Monji Kherallah1National School of Engineers of Sfax, Tunisia Faculty of Sciences of Sfax, TunisiaAbstract—In this paper, we propose a segmentation-free word recognition system based on Hidden Markov Models (HMMs) where 48 features are computed on a fixed-length sliding window, some of them are never used in literature for the recognition task. The literature has proved the difficulty of Arabic text recognition systems to recognize all character shapes which are more than 170 shapes derived from 28 basic letters. To make training and recognition of characters more efficient, we used optimized character shape models to represent the different handwritten Arabic characters. Several experiments have been performed using the IFN/ENIT database of handwritten Arabic words. The results reveal the robustness and the effectiveness of our system.
Keywords-Arabic text recognition; Hidden Markov Model; character shape models; profile based features; sliding window; IFN/ENIT database.
A new syntactic-semantic interface for ArabTAG an Arabic Tree Adjoining grammar
1Cherifa BEN KHELIL, 1Chiraz Ben Othmane Zribi, 2Denys Duchier and 3Yannick Parmentier1RIADI - ENSI, Université La Manouba, Tunisia
2Université d'Orléans, France
3Université de Lorraine, France
Abstract— For a reliable natural language processing (NLP), it is important to know how to link the meaning of a statement or a sentence to its syntactic structure. The link between semantic and syntax can be established using a syntax-semantic interface that allows the construction of sentence meaning. In this paper, we present a new approach, which built a tree adjoining grammar to represent the syntax and the semantic of modern standard Arabic. In the first part, we detail the process that automatically generates this grammar using Arab-XMG meta-grammar. Then we explain how we have established the link between syntax and semantic and how we have introduced the semantic frame -based dimension into the meta-grammar using Arabic Verbnet.
Keywords— Tree adjoining grammar; metagrammar; syntax; semantic; syntax-semantic interface ; semantic frame; Arabic language.
A Comparative Study of Filtering Approaches Applied to Color Archival Document Images
Walid Elhedda, Maroua Mehri and Mohamed Ali MahjoubLATIS Laboratory, Sousse University, National Engineering School of Sousse, Tunisia
Abstract: Current systems used by the Tunisian national archives for the automatic transcription of national archival documents
are hindered by many issues related to the performance of the optical character recognition (OCR) tools. Indeed, using
a classical OCR system to transcribe and index ancient Arabic documents is not a straightforward task due to the idiosyncrasies and the particularities of this category of documents, such as noise and degradation. Thus, applying an enhancement method or a denoising technique remains an essential prerequisite step to ease the archival document image analysis task. The stateof- the-art methods addressing the use of degraded document image enhancement and denoising are mainly based on applying filters. The most common filtering techniques applied to color images in the literature may be categorized into three classes: marginal, vector and dual approaches. A marginal approach is based on applying filters separately on each color component of the selected color space model, while a vector approach
consists in applying filters on vectors containing different color component values, determined according to the selected color
space model. A dual approach combines series and parallel filters of vector and marginal types. To provide a set of comprehensive guidelines on the strengths and the weaknesses of the most widely used marginal and the vector filtering techniques, a thorough comparative study of these techniques is proposed in this article. Numerical experiments are carried out in this study on color archival document images to show and to quantify the performance of each assessed filtering technique.
Index Terms: Historical documents, Color images, Preprocessing, Filter, Marginal approach, Vector approach.
Off-line Arabic Handwriting Recognition Using Dynamic Random Forests
Kawther Rouini, Khawla Jayech, Mohamed Ali MahjoubLATIS – Laboratory of Advanced Technology and intelligent Systems ENISO University of Sousse TunisiaAbstract: Arabic handwriting recognition is still a challenging task related especially to the unlimited variation in human handwriting, to the presence of overlapping and ligature between characters and to the high variety of Arabic character shapes. In this paper, we present an offline Arabic handwriting recognition system based on Random Forests (RFs). In fact, the RFs are a successful technique for classification and have a lot of advantages compared to other classifiers proposed in the literature to wit (i) the very high classification and recognition accuracy, (ii) the ability to determine the variable importance, and (iii) the flexibility to perform several types of statistical data analysis, including regression, classification, survival analysis, and unsupervised learning. In this study, we put forward an approach using the Surf descriptor to extract reliable features and a new RF induction algorithm called the Dynamic RFs (DRFs), which has the advantage of efficiently modeling the interaction among trees to determine the right prediction. The DRF is based on an adaptative tree induction procedure. The results carried out on the Tunisian city names database show that the DRF proves a significant improvement in terms of accuracy compared to the standard static RF and it reduces the computational time as well.
Keywords: Dynamic Random Forests, Arabic handwriting recognition, Surf descriptor.
Random Forests for the Recognition of Handwritten Arabic Mathematical Symbols
Ibtissem Haj Ali and Mohamed Ali MahjoubLATIS – Laboratory of Advanced Technology and intelligent Systems ENISO – University of Sousse Tunisia
Abstract Mathematics has a number of characteristics which distinguish it from conventional text and make it a challenging area for recognition. This include principally its two dimensional structure and the diversity of used symbols, especially in Arabic context. Recognition of mathematical formulas requires solving of three sub problems: segmentation, the symbol recognition and Finally the third step is the symbol arrangement analysis. In this paper we will focus on the symbol recognition step and we will propose a novel technique for Arabic handwritten Mathematical Symbols recognition. We constructed an invariant and efficient feature set by combination of global and local directional Chain Code Histogram (CCH) and Histogram of Oriented Gradient (HOG). For classification phase, Random Forests with Dynamic version for the induction of the forest was used. The system was evaluated on HAMF handwritten Arabic mathematical dataset. The experimental results represent 94,78 % classification rate in the test set and 99.98% in the train set.
Keywords— Arabic handwritten Mathematical Symbols recognition; HOG; CCH; Random Forests.
Secure and Energy-effictive CoAP Application Layer Protocol for the Internet of Things
1FirasAlbalas, 2Omar Almomani, 3Majd Al-Soud, 4Ammar Almomani1,3Department of Computer Science, Jordan University of Science and Technology, Jordan2Network Computer and Information Systems Department, The World Islamic Sciences & Education University, Jordan4Department of Information Technology, Al-Huson University College/ Al-Balqa Applied University, JordanAbstract—Nowadays, the concept of the Internet of Things (IoT) has become more noticeable and realistic where it is being used in all aspects of life, such as smart cities, agriculture, military surveillance, home automation, security, healthcare, etc. The idea behind IoT is making everyday objects, especially those with constrained resources, to communicate effectively with each other and with the internet; making it possible for anyone to communicate with both digital and physical worlds at any time. To enable the communication of constrained devices the Internet Engineering Task Force (IETF) came up with a Constrained Application Protocol (CoAP) to be used in the IoT application layer as a replacement for the HTTP protocol. However, the heterogeneity of the constrained devices and the complexity of the internet bring up the need for a security system to secure all the communications, data and participating things. In this paper, we proposed a lightweight secure CoAP using Elliptic Curve Cryptography (ECC) to transport security between IoT objects and the resource directory (RD). The advantage of using ECC is its compact key size enabling it to utilize a smaller key size compared to the other identical methods such as RSA. This, in turn, increases the computational speed and at the same time saves more energy. This paper mainly concentrates on comparing CoAP using the ECC method to CoAP using the Rivest–Shamir–Adleman (RSA) Cryptography algorithm in terms of the energy consumption.
Keywords— Internet of Things;CoAP; ECC; Energy saving, security, IoT.
Aspect-Based Sentiment Analysis of Arabic Laptop Reviews
Mahmoud Al-Ayyoub,1 Amal Gigieh,2 Areej Al-Qwaqenah,2 Mohammed N. Al-Kabi,3 Bashar Talafhah,1 and Izzat Alsmadi41 Computer Science Department, Jordan University of Science and Technology, Irbid, Jordan2 Computer Department, Ajloun College, Al-Balqa' Applied University, Ajloun, Jordan3 Information Technology Department, Al-Buraimi University College, Buraimi, Oman4 Department of Computing and Cyber Security, Texas A&M University-San Antonio, San Antonio, TX 78224, USAAbstract: Sentiment Analysis (SA) is one of the hottest research areas in Natural Language Processing (NLP) with vast commercial as well as academic applications. One of the most interesting versions of SA is called Aspect-Based SA (ABSA). Currently, most of the researchers focus on English text. Other languages such as Arabic have received less attention. To the best of our knowledge, only few papers have addressed ABSA of Arabic reviews and they have all been applied on only three datasets. In this work, we demonstrate our efforts to build the Arabic Laptops Reviews (ALR) dataset, which focuses on laptops reviews written in Arabic. To make it easy to use, the ALR dataset is prepared according to the annotation scheme of SemEval16-Task5. The annotation scheme considers two problems: aspect category prediction and sentiment polarity label prediction. It also comes with an evaluation procedure that extracts n-grams' features and employs a Support Vector Machine (SVM) classifier in order to allow researchers to gauge and compare the performance of their systems. The evaluation results show that there is a lot of room for improvements in the performance of the SVM classifier for the aspect category prediction problem. As for the sentiment polarity label prediction, SVM's accuracy is actually high.
Keywords: Aspect Based Sentiment Analysis; Arabic Laptop Reviews; Aspect Category Prediction; Sentiment Polarity Label Prediction; N-Gram Features; SVM Classifier;
Arabic Question Classification using Machine Learning Approaches
Imane Lahbari, Said El Alaoui Ouatik and Khalid Alaoui ZidaniSidi Mohamed Ben Abdellah University, MoroccoAbstract—Question Classification is a very important component in question answering system. In this paper, we present a machine learning comparison study for classifying Arabic questions. We have used two taxonomies: Arabic taxonomy and Li & Roth taxonomy. We have conducted several experiments using TREC and CLEF questions.
Keywords— Natural Language processing; Arabic Question Answering System; Question Classification; Taxonomy; Machine learning Approach; SVM; Decision-tree; Naive Baye.
Automatically Determining Correct Application of Basic Quranic Recitation Rules
Nour Alhuda Damer, Mahmoud Al-Ayyoub, and Ismail HmeidiJordan University of Science and Technology, Irbid, Jordan
Abstract—Quranic Recitation Rules (Ahkam Al-Tajweed) are the articulation rules that should be applied properly when
reciting the Holy Quran. Most of the current automatic Quran recitation systems focus on the basic aspects of recitation, which
are concerned with the correct pronunciation of words and neglect the advanced Ahkam Al-Tajweed that are related to the rhythmic and melodious way of recitation such as where to stop and how to "stretch" or "merge" certain letters. The only existing works on the latter parts are limited in terms of the rules they consider or the parts of Quran they cover. This paper comes to fill these gaps. It addresses the problem of identifying the correct usage of Ahkam Al-Tajweed in the entire Quran. Specifically, we focus on eight Ahkam Al-Tajweed faced by early learners of recitation. Popular audio processing techniques for feature extraction (such as LPC, MFCC and WPD) and classification (KNN, SVM, RF, etc.) are tested on an in-house dataset. Moreover, we study the significance of the features by performing several t-tests. Our results show the highest accuracy
achieved is 94.4%, which is obtained when bagging is applied to SVM with all features except for the LPC features.
Critical Proficiencies for Requirements Analysts: Reflect a Real-world needs
Issam Jebreen and Ahmad Al-QeremFaculty of Information Technology Zarqa UniversityAbstract Requirements determination (RD) is regarded as a critical phase of software development, In particularly the involve of human interaction with RD diversity increase of communication issues such as miscommunication, misunderstandings between stakeholders that impact on software projects time and cost. Therefore, the software analysts‟ communication skills are a key factor in project success. Originally analysts‟ responsibility is RD tasks, however, due to the variety and the number of tasks that need to be covered, as well as different skills for each task, the sphere of their job is usually extended. This study is explored analysts‟ proficiencies in requirement determination. An Ethnography method has been used with software Development Company in order to investigate the analysts‟ proficiencies. Our research design conducted thorough an interpretive philosophy using thematic analysis data driven approach. We have found that 18 critical proficiencies are impacting by situations in which requirement determination occurs. We propose that the analysts‟ proficiencies are a set of activities between analysts and users in which requirement determination situations consists of gathering users‟ initial requirements follow by deeply understanding of the users‟ requirements. Surprisingly, knowledge of requirements analysis and design solution methodologies including the traditional approach did not seem to be critical proficiencies for requirements analysts. In other hand, knowledge of commercial software and business process for various types of commercial business seem to be one of the most important critical proficiencies for requirements analysts.
Keywords Requirements determination, Analysts‟ proficiencies.
Enhanced LEACH Protocol for Wireless Sensor Networks
Amer O. Abu SalemDepartment of Computer Science Faculty of Information Technology, Zarqa University Zarqa, JordanAbstract Wireless Sensor Networks are been used in many practical application such as health care monitoring, environmental detection, industrial supervising and data logging. However, in order to reach their potential, researchers must discover solution to some troubles which are slowing down the wide spread use of these infrastructures. The power consumption in wireless sensor network is a primary concern and it is studiedwhen designing sensor networks. To achieve balanced energy consumption and scalability, the designers use hierarchical routing protocols based on clustering. In this paper, we study the designdifficultiesin the wireless sensor networks to setup the cluster and featureemployed of Low Energy Adaptive Clustering Hierarchy protocol (LEACH). The Enhanced protocol adds the distance parameter to the threshold T(n). The multihop routing algorithm of cluster head is proposedtoo, it based on the hop count. MATLAB was used to simulate, which is concluded that the enhanced LEACH protocol can balance load of the network to extend the life cycle of network.
Towards an online Emotional Recognition System for Intelligent Tutoring Environment
Nouha Khediri, Mohamed Ben Ammar and 1Monji Kherallah
Northern Border University, Faculty of Computing and IT, KSA
1Faculty of Sciences, University of Sfax, Tunisia
Abstract: Emotion recognition can be used in wide range of applications, we are interested in E-learning system because of
several benefits of learning anywhere anyplace and at anytime .The vision of affective computing is to impart to system, the possibility of understanding the user through recognizing his emotions and expressions. Obviously, the entire emotional state of a user is expressed and can be observed in different modalities. In this paper, a survey was conducted on the latest research in emotion recognition, first with unimodal system then with multimodal system, specifically applied in E-learning environment. Then, we propose the improvement of EMASPEL (Emotional Multi-Agents System for Peer-to-peer E-Learning) by the fusion of several realtime modalities of emotional communication.
Index Terms: Emotion Recognition, E-learning, Affective Computing, Fusion, Intelligent Tutoring System.
Multi-camera 3D modeling of a human body using the Shape from Silhouettes technique
Ahlem Youssef and Sami GAZZAH LATIS- Laboratory of Advanced Technology and Intelligent Systems, ENISo, Sousse University - TunisiaAbstract—3D modeling is becoming increasingly popular in many areas such as computer-aided design, biomedical imaging, video games, 3D reconstruction of real objects and 3D special effects. The domain of three-dimensional reconstruction is a subject of several approaches, which can be classified into three main categories: mono, stereo and multi reconstruction. In this paper, we are interested in the reconstruction of 3D multi-camera and in particularly the technique Shape from Silhouette (SFS). The main objective is the 3D modeling of a dynamic human body by a multi-camera vision system using the Shape from Silhouette technique. The first task, our system takes captures of the images from the video sequences by the OpenCV library and then it segments each image using the technique of segmentation by thresholding. To show the camera synchronization, the mapping is used to establish a point-to-point relationship between two images. Our contribution is made to the level of point of detection. The second task of our system is the 3D reconstruction from the images coming from the several cameras of a published database.
Keywords—3D reconstruction, multi-sources, multi-camera calibration, SFS, 3D model.
Vehicle Re-identification in Camera Networks: A Review and New Perspectives
Sami GAZZAH and Najoua Essoukri Ben AmaraLATIS- Laboratory of Advanced Technology and Intelligent Systems,
ENISo, Sousse University - Tunisia
Abstract—Vehicle re-identification aims to establish a match of the same vehicle in traffic-scenes over different cameras. Moving vehicle often changes appearance when located at different physical locations or regions duo to the lighting and angle of view variations. This paper provides a survey of the recent progress in the literature of vehicle re-identification in camera networks. The survey highlights the most recent approaches used to establish the correspondence between different video sequences provided by different cameras. The vehicles re- identification stills not a well investigated issue compared to the person re-identification. Several key issues still uncovered in the literature are well-covered. The paper points to several promoting directions for future research.
Keywords— Vehicle re-identification; field-of-view; Vehicle detection; styling; insert (key words).
Tracking of multiple objects based on the histogram of oriented gradients for pedestrian detection
Amari Dakhlia and Sami GAZZAHLATIS- Laboratory of Advanced Technology and Intelligent Systems,
ENISo, Sousse University - Tunisia
Abstract—Thiswork presents our detection system and tracking pedestrian-based tracklets in a complex scene. The first task of our system is the detection of pedestrians using the HOG detector, this system is able to recognize all the pedestrians present in the scene. The second task is the tracking of all detected pedestrians to get complete trajectories for a long time while preserving the identity of each pedestrian, for this we used the notion of tracklets. The experiments show that our approach outperforms the detector to find the undetected objects and the developed method eliminates the false positives and shows the effectiveness of tracking.
Keywords— Multi-objects tracking, Tracklets, HOG, Occlusion.
Multi-Gene Genetic Programming for predict rainfall data
Mohammed Alweshah, Mohammed Ababneh and 1Almahdi AlshareefAl-Balqa Applied University, Jordan 1Sabha University Sabha, Libyan Arab JamahiriyaAbstract Since the beginning of humanity, people have been interested in weather and environment such rainfall, humidity and temperature. A lot of environment specialist has tried to find a lot of methods that would allow them to know the weather in advance. This paper addresses the task of prediction rainfall data by depending on historical data and applying genetic algorithm. The quality of the data is the most important factor influence the performance of the genetic algorithm. The goal of this work is to predict the future rainfall amounts of the selected city as these predictions are very helpful and because rainfall prediction is very important for a lot of sectors such commercial, industrial, tourism and academic purposes. In this work, the genetic algorithm model is suitable for such purposes of predictions.
Keywords: Optimization, rainfall data, classification problem, genetic programming model.
Smart Real-Time RC Circuit Remote Laboratory Architecture
Fahd Ouatik1, Mustapha Raoufi1 , Mohamed Skouri1 ,Belaid Bouikhalene21Department of Physics, Laboratory of Physics, High Energy and Astrophysics Cadi Ayyad University, Morocco2Department of Computers Sciences, Sultan Moulay Slimane University, Beni Mellal, MoroccoAbstract—The smart remote laboratory experiments in engineering education became a useful tool as a great challenge for specialists. For to solve the problem of Expensive equipment and usually there are not enough devices or time for conducting experiments in a real lab. Other factors that prevent the use of lab devices directly by students are inaccessible or dangerous phenomena, or polluting chemical reactions. Setting up a new educational platform with remote laboratory experiments, many students from many countries can access them by web in order to complete, enhance their education in engineering . An advanced software/hardware flexible and real-time Operational Amplifier architecture brings additional strategies of learning and teaching, is presented in this paper. The software part is based on the html5,Canvas,javascript,Nodejs and the hardware : pcduino ,electronic device , agilent oscilloscope.
Keywords— Remote Laboratory; operational Amplifier ; Rea-time ; Node js ; HTML5; javascript; socket ; canvas; pcduino ; linux ; e-learning; Agillent technology.
Authentication Enhancement Using Mobile-Based Application
Ibrahim Fadul Ibrahim Osman and 1Yasir Abdelgadir MohamedCollege of Ahfad for Women, Ahfad University, Sudan1Faculty of Computer Science & IT, Karary University, SudanAbstract—Information security is one of the most complicated and difficult issues to which all interested parties attach a great deal of attention. Security and confidentiality studies of information, therefore, prevail over studies in other fields. The loss of information and the destruction of it by falling into hands of the abusers have many images and different forms, all of which focus on circumventing the regulations and entering them illegally. In this paper, an application Android relies on the many advantages of the Android system to authenticate a user on the mobile phone, so that they can login to a system that built, installed on the environment of the computer. The application is designed depending on the features of the mobile operating system, in the secure transport and control of data transmission, in addition to the features of the mobile phone itself, thus, features integrated to raise the level of protection of information to increase the strength authenticate to a higher extent.
Keywords—infosec; confidentiality; integrity; availability; authentication; access resource.
Large-Scale Arabic Text Classification Using MapReduce
Maher M. Abushab, Rebhi S. BarakaFaculty of Information Technology, Islamic University of Gaza, PalestineAbstract:Text classification on large-scale real documents has become one of the most core problems in text mining. For English and other languages many text classification works have been done with high performance. However, Arabic language still needs more attention and research since it is highly rich and requires special processing. Existing Arabic text classification approaches use techniques such as feature selection, data representation, feature extraction and sequential algorithms. Few attempts were done to classify large-scale Arabic text document in a parallel manner.This paper presents a parallel classification approach based on the Naïve Bayes algorithm for large volume of Arabic text using MapReduce with enhanced speedup, and preserved accuracy.The experiments show that the parallel classification approach can process large volume of Arabic text efficiently on a MapReduce and can significantly improve the speedup. Also, classification results show that the parallel classifier has achieved accuracy close to 97%.
Keywords:Text Classification, Naïve Bayes Algorithm, Parallel Classifier, MapReduce, and Hadoop.