Semantic Similarity Analysis for Corpus Development and Paraphrase Detection in Arabic

Adnen Mahmoud^1,2, and Mounir Zrigui¹

¹University of Monastir, Research Laboratory in Algebra, Numbers Theory and Intelligent Systems RLANTIS, Tunisia

²University of Sousse, Higher Institute of Computer Science and Communication Techniques ISITCom, Tunisia

Abstract: Paraphrase detection allows determining how original and suspect documents convey the same meaning. It has attracted attention from researchers in many Natural Language Processing (NLP) tasks such as plagiarism detection, question answering, information retrieval, etc., Traditional methods (e.g., Term Frequency-Inverse Document Frequency (TF-IDF), Latent Dirichlet Allocation (LDA), and Latent Semantic Analysis (LSA)) cannot capture efficiently hidden semantic relations when sentences may not contain any common words or the co-occurrence of words is rarely present. Therefore, we proposed a deep learning model based on Global Word embedding (GloVe) and Recurrent Convolutional Neural Network (RCNN). It was efficient for capturing more contextual dependencies between words vectors with precise semantic meanings. Seeing the lack of resources in Arabic language publicly available, we developed a paraphrased corpus automatically. It preserved syntactic and semantic structures of Arabic sentences using word2vec model and Part-Of-Speech (POS) annotation. Overall experiments shown that our proposed model outperformed the state-of-the-art methods in terms of precision and recall.

Keywords: Arabic language processing, word2vec, part-of-speech annotation, paraphrasing, semantic analysis, recurrent convolutional neural networks.

Received January 24, 2019; accepted February 5, 2020

https://doi.org/10.34028/iajit/18/1/1

Full Text

Speech Synthesis System for the Holy Quran Recitation

Nadjla Bettayeb and Mhania Guerti

Department of Electronics, Signal and Communications Laboratory Ecole Nationale Polytechnique, Algeria

Abstract: This paper aims to develop a Text-To-Speech (TTS) synthesis system for the holly Quran recitation, to properly helps reciters and facilitates its use. In this work, the unit selection method is adopted and improved to reach a good speech quality. The proposed approach consists mainly of two steps. In the first one, an Expert System (ES) module is integrated by employing Arabic, Quran language, phonetic and phonological features. This part was considered as a preselection to optimize the synthesis algorithm's speed. The second step is the final selection of units by minimizing a concatenation cost function and a forward-backward dynamic programming search. The system is evaluated by native and non-native Arabic speakers. The results show that the goal of a correct Quran recitation by respecting its reading rules was reached, with 97 % of speech intelligibility and 72.13% of naturalness.

Keywords: Speech synthesis, holy Quran, unit selection, expert system, Arabic language processing, tajweed rules.

Received March 21, 2019; accepted April 10, 2020

A Distributed Framework of Autonomous Drones

for Planning and Execution of Relief Operations

during Flood Situations

Zobia Zafar¹, Muhammad Awais¹, Abdul Jaleel¹, and Fiaz Majeed²

¹Department of Computer Science, University of Engineering and Technology, Pakistan

²Department of Software Engineering, University of Gujrat, Pakistan

Abstract: Every year, flood hits the world economy by billions of dollars, costs thousands of human and animal lives, destroys a vast area of land and crops, and displaces large populations from their homes. The flood affected require a time-critical help, and a delay may cause the loss of precious human lives. The ground rescue operations are difficult to carry out because of the unavailability of transport infrastructure. However, drones, Unmanned Vehicles, can easily navigate to the areas where road networks have been destroyed or become ineffective. The fleet participating in the rescue operation should have drones with different capabilities in order to make the efforts more successful. A majority of existing systems in the literature offered a centralized system for these drones. However, the performance of the existing system starts decreasing as the required number of tasks increases. This research is based on the hypothesis that a distributed intelligent method is more effective than the centralized technique for relief operations performed by multiple drones. The research aims to propose a distributed method that allows a fleet of drones with diverse capabilities to communicate and collaborate, so that the task completion rate of rescue operations could be increased. The proposed solution consists of three main modules: 1) communication and message transmission module that enables collaboration between drones, 2) realignment module that allows drones to negotiate and occupy the best position in the air to optimize the coverage area, 3) situation monitoring module that identifies the ground situation and acts accordingly. To validate the proposed solution, we have performed a simulation using AirSim simulator and compared the results with the centralized system. The proposed distributed method performed better than legacy systems. In the future, the work can be extended using reinforcement learning and other intelligent algorithms.

Keywords: Autonomous drones, flood relief operations, distributed systems, artificial intelligence, distributed collaboration.

Received October 11, 2019; accepted July 14, 2020

Lean Database: An Interdisciplinary Perspective Combining Lean Thinking and Technology

Jamil Razmak¹, Samir Al-Janabi², Faten Kharbat³, and Charles Bélanger⁴

¹College of Business, Al Ain University, UAE

²Department of Computing and Software, McMaster University, Hamilton, Canada

³College of Engineering, Al Ain University, UAE

⁴Faculty of Management, Laurentian University, Canada

Abstract: The continuous improvement approach is key to achieve a sustainable competitive advantage for organizations in their business processes. Nowadays, organizational business processes are seen through an automated function under the umbrella of organizational information systems. The huge amount of automated business processes produces data embedded with a part of messy data that could provide corrupt data. This study uses a lean thinking concept integrated with the data cleaning approach to reduce the waste of data according to business requirements and to enhance continuous improvement as part of a data defect reduction strategy. A new approach of improving and cleaning data waste is proposed by combining data cleaning algorithm and lean thinking concepts. After testing the quality and scalability of the algorithm, along with the evaluation of a corrupt dataset, the results showed improvement in the corrupt dataset reduction, leading to higher organizational performance in business processes. This integration can help researchers and technologists to fully understand and benefit from interdisciplinary capabilities while building bridges between different fields.

Keywords: Lean database, interdisciplinary, lean thinking, data quality, data cleaning.

Received March 3, 2020; accepted July 14, 2020

Reliability-Aware: Task Scheduling in Cloud Computing Using Multi-Agent Reinforcement Learning Algorithm and Neural Fitted Q

Husamelddin Balla, Chen Sheng, and Jing Weipeng

College of Information and Computer Engineering, Northeast Forestry University, China

Abstract: Cloud computing becomes the basic alternative platform for the most users application in the recent years. The complexity increasing in cloud environment due to the continuous development of resources and applications needs a concentrated integrated fault tolerance approach to provide the quality of service. Focusing on reliability enhancement in an environment with dynamic changes such as cloud environment, we developed a multi-agent scheduler using Reinforcement Learning (RL) algorithm and Neural Fitted Q (NFQ) to effectively schedule the user requests. Our approach considers the queue buffer size for each resource by implementing the queue theory to design a queue model in a way that each scheduler agent has its own queue which receives the user requests from the global queue. A central learning agent responsible of learning the output of the scheduler agents and direct those scheduler agents through the feedback claimed from the previous step. The dynamicity problem in cloud environment is managed in our system by employing neural network which supports the reinforcement learning algorithm through a specified function. The numerical result demonstrated an efficiency of our proposed approach and enhanced the reliability.

Keywords: Reinforcement learning, multi-agent scheduler, neural fitted Q, reliability, cloud computing, queuing theory.

Received April 5, 2018; accepted January 28, 2020

Formulation of Two-Stage Problem of Structural-Parametric Synthesis of Adaptive Electronic Document Management System

Artem Obukhov¹, Mikhail Krasnyanskiy², and Denis Dedov³

¹Department of Automated Decision Support Systems, Tambov State Technical University, Russian Federation

²Department of Administration, Tambov State Technical University, Russian Federation

³Department of Science, Tambov State Technical University, Russian Federation

Abstract: The paper scrutinizers the options of optimization and adaption to the individual characteristics of users of Electronic Document Management Systems (EDMS). The problem solution requires further development of the necessary methods, models and criteria. Therefore, the article considers the problem of the structural-parametric synthesis of adaptive EDMS. On the basis of previously conducted research in this area, a new architecture of EDMS is developed, within which a mathematical model of adaptive EDMS is proposed. It includes the main components of the information system, as well as a set of estimates for a number of indicators: total discounted costs, productivity, software quality and, above all, adaptability to the requirements of the user. Using this mathematical model, a two-stage task of structural-parametric synthesis of an adaptive EDMS was set, at the first stage of which the system is synthesized according to the criterion of economic efficiency, and at the second stage the process happens according to its adaptation to each user. The scientific novelty of the proposed approach consists in dividing the optimization task into two stages such as the formalization of the criteria for adaptive EDMS and development of a new architecture and mathematical model of adaptive EDMS. The results can be used to solve problems of design, modernization and adaptation of various information systems.

Keywords: Adaptability, electronic document management systems, optimization problem statement, structural-parametric synthesis.

Received May 6, 2019; accepted May 4, 2020

Secured Data Storage and Retrieval using Elliptic

Curve Cryptography in Cloud

Pradeep Suthanthiramani¹, Muthurajkumar Sannasy ², Ganapathy Sannasi ³, and Kannan Arputharaj¹

¹Department of Information Science and Technology, Anna University, India

²Department of Computer Technology, Anna University, India

³Research Centre for Cyber-Physical Systems and School of Computer Science and Engineering, Vellore Institute of Technology, India

Abstract: Security of data stored in the cloud databases is a challenging and complex issue to be addressed due to the presence of malicious attacks, data breaches and unsecured access points. In the past, many researchers proposed security mechanisms including access control, intrusion detection and prevention models, Encryption based storage methods and key management schemes. However, the role based access control policies that were developed to provide security for the data stored in cloud databases based on the sensitivity of the information are compromised by the attackers through the misuse of privileges gained by them from multiple roles. Therefore, it is necessary to propose more efficient mechanisms for securing the sensitive information through attribute based encryption by analyzing the association between the various attributes. For handling the security issue related to the large volume of cloud data effectively, the association rule mining algorithm has been extended with temporal constraints in this work in order to find the association among the attributes so that it is possible to form groups among the attributes as public attributes with insensitive data, group attributes with medium sensitive data and owner with highly sensitive attributes and data for enhancing the strength of attribute based encryption scheme. Based on the associations among the attributes and temporal constraints, it is possible to encrypt the sensitive data with stronger keys and algorithms. Hence, a new key generation and encryption algorithm is proposed in this paper by combining the Greatest common divisor and the Least common multiple between the primary key value and the first numeric non key attribute that is medium sensitive attributes and data present in the cloud database for providing secured storage through effective attribute based encryption. Moreover, a new intelligent algorithm called Elliptic Curve Cryptography with Base100 Table algorithm is also proposed in this paper for performing encryption and decryption operations over the most sensitive data for the data owners. From the experiments conducted in this work, it is observed that the proposed model enhances the data security by more than 5% when it is compared with other existing secured storage models available for cloud.

Keywords: Cloud database, secured storage, association rule mining, greatest common divisor, least common multiple, key generation and encryption.

Received July 19, 2019; accepted June 17, 2020

Middle Eastern and North African English Speech

Corpus (MENAESC): Automatic Identification

of MENA English Accents

Abstract: This study aims to explore the English accents in the Arab world. Although there are limited resources for a speech corpus that attempts to automatically identify the degree of accent patterns of an Arabic speaker of English, there is no speech corpus specialized for Arabic speakers of English in the Middle East and North Africa (MENA). To that end, different samples were collected in order to create the linguistic resource that we called Middle Eastern and North African English Speech Corpus (MENAESC). In addition to the “accent approach” applied in the field of automatic language/dialect recognition; we applied also the “macro-accent approach” -by employing Mel-Frequency Cepstral Coefficients (MFCC), Energy and Shifted Delta Cepstra (SDC) features and Gaussian Mixture Model-Universal Background Model (GMM-UBM) classifier- on four accents (Egyptian, Qatari, Syrian, and Tunisian accents) among the eleven accents that were selected based on their high population density in the location where the experiments were carried out. By using the Equal Error Rate percentage (EER%) for the assessment of our system effectiveness in the identification of MENA English accents using the two approaches mentioned above through the employ of the MENAESC, results showed we reached 1.5 to 2%, for “accent approach” and 2 to 3.5% for “macro-accents approach” for identification of MENA English. It also exhibited that the Qatari accent, of the 4 accents included, scored the lowest EER% for all tests performed. Taken together, the system effectiveness is not only affected by the approaches used, but also by the database size MENAESC and its characteristics. Moreover, it is impacted by the proficiency of the Arabic speakers of English and the influence of their mother tongue.

Keywords: MENAESC, MFCC+Energy and SDC features, accent, macro-accent, automatic identification.

Parallel Scalable Approximate Matching Algorithm

for Network Intrusion Detection Systems

Adnan Hnaif¹, Khalid Jaber¹, Mohammad Alia¹, and Mohammed Daghbosheh²

¹Faculty of Science and Information Technology, Al Zaytoonah University of Jordan, Jordan

²Faculty of Science and Information Technology, Irbid National University of Jordan, Jordan

Abstract: Matching algorithms are working to find the exact or the approximate matching between text “T” and pattern “P”, due to the development of a computer processor, which currently contains a set of multi-cores, multitasks can be performed simultaneously. This technology makes these algorithms work in parallel to improve their speed matching performance. Several exact string matching and approximate matching algorithms have been developed to work in parallel to find the correspondence between text “T” and pattern “P”. This paper proposed two models: First, parallelized the Direct Matching Algorithm (PDMA) in multi-cores architecture using OpenMP technology. Second, the PDMA implemented in Network Intrusion Detection Systems (NIDS) to enhance the speed of the NIDS detection engine. The PDMA can be achieved more than 19.7% in parallel processing time compared with sequential matching processing. In addition, the performance of the NIDS detection engine improved for more than 8% compared to the current SNORT-NIDS detection engine.

Keywords: Exact matching algorithms, approximate matching algorithms, parallel processing, network intrusion detection systems.

Received February 13, 2020; accepted June 17, 2020

Instagram Post Popularity Trend Analysis and Prediction using Hashtag, Image Assessment, and User History Features

Kristo Radion Purba, David Asirvatham, and Raja Kumar Murugesan

School of Computer Science and Engineering, Taylor's University, Malaysia

Abstract: Instagram is one of the most popular social networks for marketing. Predicting the popularity of a post on Instagram is important to determine the influence of a user for marketing purposes. There were studies on popularity prediction on Instagram using various features and datasets. However, they haven't fully addressed the challenge of data variability of the global dataset, where they either used local datasets or discretized output. This research compared several regression techniques to predict the Engagement Rate (ER) of posts using a global dataset. The prediction model, coupled with the results of the popularity trend analysis, will have more utility for a larger audience compared to existing studies. The features were extracted from hashtags, image analysis, and user history. It was found that image quality, posting time, and type of image highly impact ER. The prediction accuracy reached up to 73.1% using the Support Vector Regression (SVR), which is higher than previous studies on a global dataset. User history features were useful in the prediction since the data showed a high variability of ER if compared to a local dataset. The added manual image assessment values were also among the top predictors.

Keywords: Social media, Instagram, popularity trend, machine learning, prediction model.

Received February 17, 2020; accepted August 6, 2020

A New Image Encryption Scheme Using Dual Chaotic Map Synchronization

Obaida Al-Hazaimeh¹, Mohammad Al-Jamal², Mohammed Bawaneh¹, Nouh Alhindawi³, and Bara’a Hamdoni²

¹Department of Computer Science and Information Technology, Al- Balqa Applied University, Jordan

²Department of Mathematics, Yarmouk University, Jordan

³Faculty of Sciences and Information Technology, Jadara University, Jordan

Abstract: Chaotic systems behavior attracts many researchers in the field of image encryption. The major advantage of using chaos as the basis for developing a crypto-system is due to its sensitivity to initial conditions and parameter tunning as well as the random-like behavior which resembles the main ingredients of a good cipher namely the confusion and diffusion properties. In this article, we present a new scheme based on the synchronization of dual chaotic systems namely Lorenz and Chen chaotic systems and prove that those chaotic maps can be completely synchronized with other under suitable conditions and specific parameters that make a new addition to the chaotic based encryption systems. This addition provides a master-slave configuration that is utilized to construct the proposed dual synchronized chaos-based cipher scheme. The common security analyses are performed to validate the effectiveness of the proposed scheme. Based on all experiments and analyses, we can conclude that this scheme is secure, efficient, robust, reliable, and can be directly applied successfully for many practical security applications in insecure network channels such as the Internet.

Keywords: Chaos, lorenz systems, chen systems, synchronization, cryptography.

Received April 7, 2020; accepted August 26, 2020

A Novel Approach to Maximize G-mean in Nonstationary Data with Recurrent Imbalance Shifts

Radhika Kulkarni¹, S. Revathy¹, and Suhas Patil²

¹Department of Computer Science Engineering, Sathyabama Institute of Science and Technology, India

²Department of Computer Science Engineering, Bharati Vidyapeeth’s College of Engineering, India

Abstract: One of the noteworthy difficulties in the classification of nonstationary data is handling data with class imbalance. Imbalanced data possess the characteristics of having a lot of samples of one class than the other. It, thusly, results in the biased accuracy of a classifier in favour of a majority class. Streaming data may have inherent imbalance resulting from the nature of dataspace or extrinsic imbalance due to its nonstationary environment. In streaming data, timely varying class priors may lead to a shift in imbalance ratio. The researchers have contemplated ensemble learning, online learning, issue of class imbalance and cost-sensitive algorithms autonomously. They have scarcely ever tended to every one of these issues mutually to deal with imbalance shift in nonstationary data. This correspondence shows a novel methodology joining these perspectives to augment G-mean in no stationary data with Recurrent Imbalance Shifts (RIS). This research modifies the state-of-the-art boosting algorithms,1) AdaC2 to get G-mean based Online AdaC2 for Recurrent Imbalance Shifts (GOA-RIS) and AGOA-RIS (Ageing and G-mean based Online AdaC2 for Recurrent Imbalance Shifts), and 2) CSB2 to get G-mean based Online CSB2 for Recurrent Imbalance Shifts (GOC-RIS) and Ageing and G-mean based Online CSB2 for Recurrent Imbalance Shifts (AGOC-RIS). The study has empirically and statistically analysed the performances of the proposed algorithms and Online AdaC2 (OA) and Online CSB2 (OC) algorithms using benchmark datasets. The test outcomes demonstrate that the proposed algorithms globally beat the performances of OA and OC.

Keywords: Cost-sensitive algorithms, data stream classification, imbalanced data, online learning, population shift, skewed data stream.

Received March 23, 2019; accepted April 13, 2020

An Efficient Intrusion Detection System by Using Behaviour Profiling and Statistical Approach Model

Rajagopal Devarajan and Padmanabhan Rao

PG and Research Department of Computer Science and Applications, Vivekanandha College of Arts and Sciences for Women (Autonomous), India

Abstract: Unauthorized access in a personal computer or single system of a network for tracking the system access or theft the information is called attack/ hacking. An Intrusion detection System defined as an effective security technology, it detect, prevent and possibly react to computer related malicious activities. For protecting computer systems and networks from abuse used mechanism named Intrusion detection system. The aim of the study is to know the possibilities of Intrusion detection and highly efficient and effective prevent technique. Using this model identified the efficient algorithm for intrusion detection Behaviour Profiling Algorithm and to perform dynamic analysis using Statistical Approach model using log file which provides vital information about systems and the activities on them. The proposed algorithm implemented model it produced above 90%, 96% and 98% in the wired, wireless and cloud network respectively. This study concluded that, the efficient algorithm to detect the intrusion is behaviour profiling algorithm, while join with the statistical approach model, it produces efficient result. In further research, possibility to identify which programming technique used to store the activity log into the database. Next identify which algorithm is opt to implement the intrusion detection and prevention system by using big data even the network is wired, wireless or cloud network.

Keywords: IDS, IPS, behaviour profiling algorithm, statistical approach model, NIDS, HIDS.

https://doi.org/10.34028/iajit/18/1/13

Full Text  

Pain Detection/Classification Framework including Face Recognition based on the Analysis of Facial Expressions for E-Health Systems

Fatma Elgendy¹, Mahmoud Alshewimy², and Amany Sarhan²

¹Kafrelshiekh Higher Institute for Engineering and Technology, Egypt

²Computer and Control Engineering Department, Tanta University, Egypt

Abstract: Facial expressions can demonstrate the presence and degree of pain of humans, which is a vital topic in E-healthcare domain specially for elderly people or patients with special needs. This paper presents a framework for pain detection, pain classification, and face recognition using feature extraction, feature selection, and classification techniques. Pain intensity is measured by Prkachin and Solomon pain intensity scale. Experimental results showed that the proposed framework is a promising one compared with previously works. It achieves 91% accuracy in pain detection, 99.89% accuracy in face recognition, and 78%, 92%, 88% accuracy, respectively, for three levels of pain classification.

Keywords: E-health, Gabor filter, Adaboost, relieff filter, SADE, KNN.

Received January 12, 2020; accepted March 19, 2020