Nowadays, e-mail is widely becoming one of the fastest and most economical forms of communication .Thus, the e-mail is prone to be misused. One such misuse is the posting of unsolicited, unwanted e-mails known as spam or junk e-mails. This paper presents and discusses an implementation of an Anti-spam filtering system, which uses a Multi-Layer Perceptron (MLP) as a classifier and a Genetic Algorithm (GA) as a training algorithm. Standard genetic operators and advanced techniques of GA algorithm are used to train the MLP. The implemented filtering system has achieved an accuracy of about 94% to detect spam e-mails, and 89% to detect legitimate e-mails. 

 Abstract The Arabic language presents significant challenges to many natural language processing applications. The broken plu rals (BP) problem is one of these challenges especially for information retrieval applications. It is difficult to deal with Arabic broken plurals and reduce them to their associated singulars, because no obvious rules exist, and there are no standard stemming algorithms that can process them. This paper attempts to handle the problem of broken plural by de veloping a method to identify broken plurals in an unvowelised Arabic text and reducing them to their correct singular forms by incorporating the simple broken plural matching approach, with a machine translation system and an English stemmer as a new approach. A set of experiments has been conducted to evaluate the performance of the proposed method using a number of text samples extracted from a large Arabic corpus (AL-Hayat newspaper). The obtained re sults are analyzed and discussed.

 Abstract There are several stemming approaches that are applied to Arabic language, yet no a complete stemmer for this language is available. The existing stem-based stemmers for stemming Arabic text have a poor performance in terms of accuracy and error rates. In order to improve the accuracy rates of stemming, a hybrid method is proposed for stemming Arabic text to produce stems (not roots). The improvement of the accuracy of stemming will lead by necessity to the improvement of many applications very greatly, including: information retrieval, document classification, machine translation, text analysis and text compression. The proposed method integrates three different stemming techniques, including: morphological analysis, affix-removal and dictionaries. 

This paper investigates both the computation accuracy (based on the power balance error) and the cost in terms of the number of FDTD-grid cells due to the artifact rotation for a cellular handset close to the user’s head. Two study cases are simulated to assess the EM coupling of a cellular handset and a MRI-based human head model at 900 MHz; firstly, both handset and head CAD models are aligned to the FDTD-grid, secondly, handset close to a rotated head in compliance with IEEE-1528 standard. A FDTD-based platform, SEMCAD-X, is used; where conventional and interactive gridder approaches are implemented to achieve the simulations. The results show that owing to the artifact rotation, the computation error may increase up to 30%, whereas, the required number of grid cells may increase up to 25%

Abstract

Most of Arabic countries are located within the arid and semi-arid zones, which are characterised by the scarcity and the high spatial and temporary variability of rainfall .The influence of rainfall patterns on natural vegetation cover is significant and can be monitored and assessed using NDVI indices derived from remote sensing data. The literature stresses on the wide use of NDVI to study the changes in vegetation cover, because of the sensitivity of vegetation to the Red and Infra-red spectrum. In this paper the SPOT Apparent Green Cover Percentage data and MODIS Vegetation index are used to monitor the vegetation cover change in North western Libya. The aim of this study is to establish the spatial and temporal changes in vegetation cover and their relation to the rainfall pattern. The data and method: The utilized data for research included multi-temporal remotely sensed data (Spot, MODIS, and Landsat ETM), climatic data (ground stations and satellite data) and digital maps (DEM, slope, flow direction and basins). The satellite images were firstly pre-processed. This included the importing of different format images into a standard format of the ERDAS IMAGINE 9.1. Then the dataset were geometrically corrected into the WGS84 latitude/longitude geographic co-ordinate system. The study area was then subseted using a vector file representing the area boundary (AOI). The multibands were stacked to create time series, the NDVI equation was used to convert the DN values into NDVI values. 

As software systems grow, their complexity augments dramatically. In consequence their understandability and evolvability are becoming a difficult task, therefore a way to describe the architecture of software systems is becoming a must. The architecture of a software system can be described using either an architecture description language (ADL) or an object-oriented modeling language. In this article, we show how we can build a hybrid model, based on the two approaches, to describe the architecture of software systems. This approach profits from the advantages of the two approaches, i.e. the explicit presentation of functional aspects in object-oriented modeling and the explicit separation of competition (components) from interactions (connectors) in architectural description.