DSpace Community: College of Computer and Information Sciences

Secure Identification System-SIS

Title: Secure Identification System-SIS

Authors: Rahal, Dr. Salah M.; Abu Samah, Hatim A.; Muteb, Khaled N.

Abstract: The secure identification of a person was and still of great interest in differentfields of human activities. The traditional identification methods such as PIN,passwords aren’t capable to achieve a high level of secure identification. In thecontrast, the biometrics allows to attain the suitable solution for this vital question.Different types of biometrics are reviewed. Its characteristics are unique for eachperson either for physiological or behavioral biometrics. It is shown that thefingerprint recognition has a very good balance of all desirable properties.The design of secure identification fingerprint-based system, in its hardware andsoftware parts, is done. This system was implemented with the suitable features.

Image Segmentation Using Split and Merge Techniques with a Mixture of Heterogeneous Distributions

Mon, 25 Apr 2011 00:00:00 GMT

Title: Image Segmentation Using Split and Merge Techniques with a Mixture of Heterogeneous Distributions

Authors: Reyad Ali, Yaser Ali; El-Zaart, Dr. Ali

Abstract: Image segmentation is a technique that partitioned the image into prerequisite semantic unique regions. Simplifying the representation of an image into something more easily to analyze and meaningful is the ultimate goal of segmentation. It is used for locating boundaries and objects in an image such as lines, curves or object. Segmentation serves many computer applications such as pattern recognition, object recognition, automatic trafficcontrol, and many other applications. Image segmentation is considered the bottleneck in many image processing techniques.There are a vast number of segmentation techniques that is available but none of them satisfy the global properties, so it is remain challenge for researcher to find best one. Thresholding is one of the simplest and effective techniques for image segmentation. It defines a threshold value T, and the gray values less than T will be considered as a class, and those above T belong to another class. The estimation of the optimal threshold T is still the big problem for image thresholding.The aim of this thesis is to introduce a new method for image thresholding based on histogram split-merge technique with a mixture of heterogeneous distributions. The proposed method will be applied on different type of synthetic and real images; the results obtained will be compared to those of the existing methods.

Data Services Middleware for Data Sharing in Peer-to-Peer Environment

Sat, 21 May 2011 00:00:00 GMT

Title: Data Services Middleware for Data Sharing in Peer-to-Peer Environment

Authors: Beraka, Mutaz Seleam; GANNOUNI, Dr. Sofien

Abstract: The advances in network computing models evolve the way people are sharing their resources. Nowadays, users are considered as peers and as such they are able to share their processing power, their applications and their data with each other. File sharing model is one of P2P models that allow peers to share and exchange files between them. This model isn’t sufficient for current demands of sharing. It doesn’t benefit from the huge amount of data which was stored in data sources and located locally in peers’ machines. So, to take advantage of the data stored in heterogeneous and autonomous data sources that spread across the enterprise, we intend to develop in this thesis a middleware that facilitates data sharing between different users regardless of their heterogeneity. This middleware adopts a service-oriented architecture and relies on peer-to-peer computing techniques in order to provide a set of ready-made and easy to use components that allow inexperienced users to publish and advertize their own data sources as well as discover and consume those of others. Also, this middleware relies on process oriented approach to provide data service composition engine to deliver a semi-dynamic service composition. This type of service composition composes a set of data services to provide virtual data integration through a set of rich functionalities.By using such middleware, sharing data services between peers will become more easily as well as more reliable in P2P environment. It masks the heterogeneity between peers and the heterogeneity between data sources. Moreover, it meets user’s need of data availability and data integration through service composition. However, the proposed middleware will be able to use from any peer interested in the providing and consuming data services, and wants to take advantages of data integration that obtain from heterogeneous data sources.

Towards Developing Automatic Name-Entity-Recognition System for Arabic Text

Sat, 29 Jan 2011 00:00:00 GMT

Title: Towards Developing Automatic Name-Entity-Recognition System for Arabic Text

Authors: Naji, Fadl Dahan Abdu; Touir, Dr. Ameur

Abstract: Name Entity Recognition (NER) has emerged as a Natural Language Processing (NLP) technology that is effective and can provide high value to several different kinds of application such as Information Extraction (IE), Information Retrieval (IR), Question Answering (QA), text clustering, etc. NER is responsible for the identification of proper names in text and their classification as different types of named entity such as people, locations, and organizations. There are two main approaches to NER, one is based on linguistic knowledge in particular grammar rules and hence called rule-based, while the other is based on machine learning techniques.We aim in this research to build Name Entity Recognition automatic system for the Arabic language using Machine Learning (ML) approaches; these approaches have many models such as Maximum Entropy (ME), Decision Tree (DR), Support vector Machines (SVM), and Hidden Markov Model (HMM). Among these models we used HMM to build our system for the reason that it relies on the context structure. ML approaches provide us the ability to work with unrestricted domain, and to adapt the suitable machine learning with the nature and the difficulties of some characteristic of the Arabic language. The Arabic language does not exhibit differences in orthographic case; whereas the English language mixes case texts, therefore, there is some obvious clue such as initial capitalized letters to indicate the presence of a name constituent.