Elnaz Limouchi, Mohammadali Pourmina, Afshin Salajeghe, Ashkan Zehni
Abstract: Scheduling mechanisms for both uplink and downlink channels in IEEE 802.16 standard are open area for research. In this paper, we propose a weighted fair priority intra class scheduling for point to multipoint mobile WiMAX system. This method takes user’s battery charge level into account to determine weights of the users. Simulation results show that overall system throughput is improved.
Keywords : WiMAX, Scheduling; OFDMA; Battery level; Resource allocation.
Adnan M. Al-Khatib
Abstract: Money in the e-commerce network, represents information moving at the speed of light, where fraud (digital crime) within the banking and financial services happened very fast and can cost billions of dollars each year-undetected and unreported. In this paper I present a comprehensive framework that mines and detect fraudulent transactions of Card-Not-Present (CNP) in the e-payment systems with a high degree of accuracy.
Keywords : Credit Card; Card-Not-Present; Fraud Detection; Data Mining; Profiling; Accuracy; Rule.
Chanattha Thongsuk, Choochart Haruechaiyasak, Somkid Saelee
Abstract: Today Twitter, a social networking website, has become a new advertising channel to promote products and services using online social network community. In this study, we propose a solution to recommend Twitter users to follow businesses, which match their interests. Our approach is based on classification algorithms to predict user’s interests by analyzing their posts. The challenging issue is the short length characteristic of Twitter posts. With only a few available key terms in each post, classifying Twitter posts is very difficult and challenging. To alleviate this problem, we propose a technique to improve the classification performance by expanding the term features from a topic model to train the classification models. A topic model is constructed from a set of topics based on the Latent Dirichlet Allocation (LDA) algorithm. We propose two feature processing approaches: (1) feature transformation, i.e., using a set of topics as features and (2) feature expansion, i.e., appending a set of topics to a set of terms. Experimental results of multi-classification showed that the highest accuracy of 95.7% is obtained with the feature expansion technique, an improvement of 19.1% over the Bag of Words (BOW) model. In addition, we also compared between multi-classification and binary classification using feature expansion approach to build the classification models. The performance of feature expansion approach using binary classification yielded higher accuracy than the multi-classification equal to 2.3%, 3.3% and 0.4%, for airline, food and computer & technology businesses, respectively.
Keywords : Classification; topic model; Latent Dirichlet Allocation (LDA); Twitter.
Ankur Gupta, A. K. Vatsa
Abstract: Mobile ad hoc Network (MANET) has a challenging task because of the dynamic and infrastructure less nature of the network. This dynamic nature leads to the difficulty in handoff, addressing, routing and data delivering process. For getting good QoS, better flexibility, effective and efficient handoff process in this dynamic network need better handoff mechanism for avoiding any discontinuity, packet loss, delay and jitter during handoff process. Therefore, it is necessary to make fast handoff based on group mobility and policy driven approach for this network. Thus, In this paper, We propose a policy based handoff mechanism with group mobility over hierarchical cluster based architecture involving proactive and reactive handoff approaches based on policy for nodes of MANET.
Keywords : MANET (Mobile ad hoc Network); Fast Handoff; Proactive and Reactive handoff; Policy based approach; Group mobility; Cluster head.
Kamel Khoualdi, Marwan El-Haj Mahmoud
Abstract: Toys problems, such as the puzzle problem, are solved using classical artificial intelligence search algorithms Such as Breadth-first search and depth-first search. These strategies requires the generation of a graph known as the state space search that consist of the different states a problem may have. Using the above search techniques, a solution of the problem consists of a systematic exploration of the different state, starting from an initial state and moving towards a final state. This approach is time and memory consuming. In this paper, we propose a multiagent approach as an alternative to solve the puzzle problem. Each block in the puzzle is considered as a reactive agent. The paper shows how the solution is reached through the interaction of the agents.
Keywords : multiagent systems; puzzle problem; reactive agents; search methods.
Rekha Chakravarthi, C.Gomathy
Abstract: Congestion plays a vital role in degrading the performance of wireless sensor network. Thus an issue of detecting and controlling congestion becomes essential to improve the performance of the network. There are various sources for congestion like packet collision, buffer overflow, concurrent transmission etc. This paper focuses on congestion due to concurrent transmission. We have proposed an efficient protocol to detect and control congestion in a MAC. The level of congestion is measured using a metric called Depth of Congestion (DC). Based on the measured value the node effectively adapts its transmission data rate to control congestion. This technique is implemented successfully in NS-2 simulator. Finally, simulation results have demonstrated the effectiveness of our proposed protocol.
Keywords : Wireless Sensor Networks; Depth of Congestion (DC); Congestion Detection; Congestion Control.
L. Senthilvadivu, K. Duraiswamy
Abstract: The World Wide Web (WWW) is known for being a web of documents; however, little is known about the structure or growth of such a web. Search engines such as Google have transformed the way people access and use the web and have become a critical technology for finding and delivering information .In the proposed work, the tasks such as the information assimilation and retrieval have been discussed. In the information assimilation, the data synthesis have been done from multi related and heterogeneous information sources and stored hierarchically. The data can be collected from various resources under different domains. The data are available in persistence storage by using default programmatic methodology. In this paper we propose also a searching algorithm to be used in web search engines that simply relies on information that could be extracted based on user queries from multi related and heterogeneous information resources. Hierarchical results from Heterogeneous Domain, Build Positive Set and Fetch Positive Results are the most important aspects of the searching system. Keywords : Assimilation; Heterogeneous domain; persistence storage; Hierarchical results.
J. Umamaheswari, G. Radhamani
Abstract: Image classification is a most important step for image analysis. As the same in medical area especially for diagnosing the disease of the patient, classification plays a great role for the doctors to treat the patient according to the severeness of the diseases. In case of DICOM images it is very tough for optimal identification and early detection of diseases. Classification is a computational procedure that separates the images into groups according to their features that extracted. DICOM is latest medical imaging technology. DICOM is used for brain scans and it is very useful and effective technique to detect the dissimilarity in brain images. In this paper a hybrid approach is proposed for DICOM image classification. The approach consists of feature extraction and classification. The classification consists of Multi Linear Discriminent Analysis (MLDA) and Support Vector Machine (SVM). Classification is done on the base of parameter extracted by Gray Level Co-occurrence Matrix (GLCM) and histogram texture feature extraction method. The feature is selected using fuzzy rough set and Genetic Algorithm (GA). The proposed approach has high approximation capability and much faster convergence.
Keywords : Classification; Linear Discriminent Analysis (LDA); Support Vector Machine (SVM); GA, Fuzzy Rough set; GLCM; Histogram Texture feature.