IJITCS Vol. 8, No. 3, Mar. 2016
Cover page and Table of Contents: PDF (size: 187KB)
Multi-tenant databases (MTD) are aspect of computing that has become one of the revolutionary technologies of recent years. This technology has helped to discard the large-scale investments in hardware and software resources, in upgrading them regularly and also in expensive licences of application software used on in-house hosted database systems. A MTD is a way of deploying a Database as a Service (DaaS) for the convenience and benefits of tenants. This concept is good for higher scalability and flexibility but it involves a lot of technicalities. The adoption of a MTD is based on some salient factors which can be grouped into four categories for easy understanding. A survey is presented in this research that involves a focus group of thirty respondents. The result shows the degree of impact each factor has on the decision to adopt a MTD. This paper also considers these factors and develops a framework that will help prospective tenants to take an informed decision about the adoption of the concept.[...] Read more.
Separating audio sources from a convolutive mixture of signals from various independent sources is a very fascinating area in personal and professional context. The task of source separation becomes trickier when there is no idea about mixing environment and can be termed as blind audio source separation (BASS). Mixing scenario becomes more complicated when there is a difference between number of audio sources and number of recording microphones, under determined and over determined mixing. The main challenge in BASS is quality of separation and separation speed and the convergence speed gets compromised when separation techniques focused on quality of separation. This work proposed divergence algorithm designed for faster convergence speed along with good quality of separation. Experiments are performed for critically determined audio recording, where number of audio sources is equal to number of microphones and no noise component is taken into consideration. The result advocates that the modified convex divergence algorithm enhance the convergence speed by 20-22% and good quality of separation than conventional convex divergence ICA, Fast ICA, JADE.[...] Read more.
The distributed environments vary largely in their architectures, from tightly coupled cluster environment to loosely coupled Grid environment and completely uncoupled peer-to-peer environment, and thus differ in their working environments as well as performance. To meet the specific needs of these environments for data organization, replication, transfer, scheduling etc. the data management systems implement different data management models. In this paper, major data management tasks in distributed environments are identified and a taxonomy of the data management models in these environments is presented. The taxonomy is used to highlight the specific data management requirements of each environment and highlight the strengths and weakness of the implemented data management models. The taxonomy is followed by a survey of different distributed and Grid environments and the data management models they implement. The taxonomy and the survey results are used to identify the issues and challenges of data management for future exploration.[...] Read more.
At a recent time, the web has become a valuable source of online consumer review however as the number of reviews is growing in high speed. It is infeasible for user to read all reviews to make a valuable or satisfying decision because the same features, people can write it contrary words or phrases. To produce a useful summary of domain synonyms words and phrase, need to be a group into same feature group. We focus on feature-based opinion mining problem and this paper mainly studies feature based product categorization from the number of users - generated review available on the different website. First, a multi-feature segmentation method is proposed which segment multi-feature review sentences into the single feature unit. Second part of speech dictionary and context information is used to consider the irrelevant feature identification, sentiment words are used to identify the polarity of feature and finally an unsupervised clustering based product feature categorization method is proposed. Clustering is unsupervised machine learning approach that groups feature that have a high degree of similarity in a same cluster. The proposed approach provides satisfactory results and can achieve 100% average precision for clustering based product feature categorization task. This approach can be applicable to different product.[...] Read more.
Knowledge based system can be designed to solve complex medical problems. It incorporates the expert’s knowledge that has been coded into facts, rules, heuristics and procedures. Incorporation of local languages with the knowledge based system allows end-users communicate with the system in a simpler and easier way. In this study a localized knowledge based system is developed for TB disease diagnosis using Ethiopian national language. To develop the localized knowledge based system, tacit knowledge is acquired from domain experts using interviewing techniques and explicit knowledge is captured from documented sources using relevant documents analysis method. Then the acquired knowledge is modeled using decision tree structure that represents concepts and procedures involved in diagnosis of disease. Production rules are used to represent domain knowledge. The localized knowledge based system is developed using SWI Prolog version 6.4.1 programming language. Prolog supports natural language processing feature to localize the system. As a result, the system is implemented using Amharic language (the national language of Ethiopia) user interface. With Localization, users at remote areas and users who are not good in foreign languages are benefited enormously. The system is tested and evaluated to ensure that whether the performance of the system is accurate and the system is usable by physicians and patients. The average performance of the localized knowledge based system has registered 81.5%.[...] Read more.
K-Modes is an eminent algorithm for clustering data set with categorical attributes. This algorithm is famous for its simplicity and speed. The K-Modes is an extension of the K-Means algorithm for categorical data. Since K-Modes is used for categorical data so 'Simple Matching Dissimilarity' measure is used instead of Euclidean distance and the 'Modes' of clusters are used instead of 'Means'. However, one major limitation of this algorithm is dependency on prior input of number of clusters K, and sometimes it becomes practically impossible to correctly estimate the optimum number of clusters in advance. In this paper we have proposed an algorithm which will overcome this limitation while maintaining the simplicity of K-Modes algorithm.[...] Read more.
Since the granting of banking facilities in recent years has faced problems such as customer credit risk and affects the profitability directly, customer credit risk assessment has become imperative for banks and it is used to distinguish good applicants from those who will probably default on repayments. In credit risk assessment, a score is assigned to each customer then by comparing it with the cut-off point score which distinguishes two classes of the applicants, customers are classified into two credit statuses either a good or bad applicant. Regarding good performance and their ability of classification, generalization and learning patterns, Multi-layer Perceptron Neural Network model trained using various Back-Propagation (BP) algorithms considered in designing an evaluation model in this study. The BP algorithms, Levenberg-Marquardt (LM), Gradient descent, Conjugate gradient, Resilient, BFGS Quasi-newton, and One-step secant were utilized. Each of these six networks runs and trains for different numbers of neurons within their hidden layer. Mean squared error (MSE) is used as a criterion to specify optimum number of neurons in the hidden layer. The results showed that LM algorithm converges faster to the network and achieves better performance than the other algorithms. At last, by comparing classification performance of neural network with a number of classification algorithms such as Logistic Regression and Decision Tree, the neural network model outperformed the others in customer credit risk assessment. In credit models, because the cost that Type II error rate imposes to the model is too high, therefore, Receiver Operating Characteristic curve is used to find appropriate cut-off point for a model that in addition to high Accuracy, has lower Type II error rate.[...] Read more.
Ranking fuzzy numbers has become an important process in decision making. Many ranking methods have been proposed thus far and one of the commonly used is centroid of trapezoid. Here we try to derive detail mathematical derivation of centroids of a Trapezoidal Intuitionistic Fuzzy Number along x and y axis. After that we derive the ranking value from two centroid along x and y axis. At the end of the article ranking value on fuzzy geometric programming is used. Here we are dealing with three strong decision making concepts. Intuitionistic trapezoidal fuzzy system is much more decision oriented approach than normal fuzzy number in real life uncertain environment, where we can apply membership and non membership concept for analyzing any real life situation. Ranking value, based on centroid of any Trapezoidal Intuitionistic Fuzzy Number helps for conclusion derivation in quantitative way. We here choose most powerful non linear optimization tool, geometrical programming technique, for generating any decision, using Trapezoidal Intuitionistic Fuzzy Number with centroid ranking approach.[...] Read more.
The application of human immunology in solving security problems in Grid Computing seems to be a thought-provoking research area. Grid involves large number of dynamic heterogeneous resources. Manually managing the security for such dynamic system is always fault prone. This paper presents the simple immune based model for self-protection (SIMS) of grid environment from various attacks like DoS, DDoS, Probing, etc. Like human body helps to identify and respond to harmful pathogens that it doesn't recognize as "self", in the same manner SIMS incorporates the immunological concepts and principles for safeguarding the grid from various security breaches.[...] Read more.
Effective design of online shopping websites is the need of the hour as design plays a crucial role in the success of online shopping businesses. Recently, the use of Quality Function Deployment (QFD) has been reported for the design of online shopping websites. QFD is a customer driven process that encompasses voluminous data gathered from customers through several techniques like personal interview, focus groups, surveys etc. This massive, unsorted and unstructured data is required to be transformed into a limited number of structured information to represent the actual Customer Needs (CNs) which are then utilized in subsequent stages of QFD process. This can be achieved through brainstorming using techniques like Affinity Process. However, integrating the Affinity Process within QFD is tedious and time consuming and cannot be dealt with manually. This generates a pressing need for a software tool to serve the purpose. Moreover, the researches carried out so far have focused on QFD application, post the generation of CNs. Also, the available QFD softwares lack the option to generate CNs from collected data. Thus, the paper aims to develop a novel software tool that integrates Affinity Process with QFD to generate customers' needs for effective design of online shopping websites. The software system is developed using Visual Basic Dot Net (VB.Net) that integrates a MS-Access database.[...] Read more.