Show: 25 50 75 100 Results

Search results: 25 out of 1,990

ازالة الضوضاء من الصور باستخدام تقنية الانصهار معتمدا على معاملات تحويل المويجة المتقطعة == Image Denoising Using Fusion Technique Based on DWT Coefficient

Author name: رسل حيدر جاسم
Supervisor name: اسراء هادي علي الشمري
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Babylon
First pages:
Abstract: الضوضاء هو عبارة عن معلومات غير مرغوب فيها في الصورة التي تدهورت اثناء عملية النقل او استحصال الصورة. ان اثار الضوضاء يقلل من وضوح الصورة. عملية ازالة الضوضاء من الصورة يسهل من عملية المعالجة. ان مشكلة ازالة الضوضاء في المجال المكاني تعمل على تنعيم البيانات بالاضافة الى تنعيم الحواف. لهذا السبب استخدمت ازالة الضوضاء عن طريق المويجة باستخدام المجال الترددي التي تعمل في المحافظة على الخصائص المهمة للصورة والحواف بالاضافة الى ازالة الضوضاء من الصور لكن تبقى مشكلة فقدان بعض المعلومات المهمة للصورة وتكوين صورة اقل وضوحا. تلعب طريقة انصهار الصور في الوقت الحاضر التي هي احدى انواع تكنولوجيا المعلومات المتكاملة دورا هاما في العديد من المجالات وتكوين صورة ذات نوعية عالية . الهدف من انصهار الصورة هو مزج المعلومات من صور عديدة بالاضافة الى حفظ المعلومات الهامة الموجودة في الصور الاصلية واستعادة جودة الصورة بناءا على مقياس نسبة الاشارة الى الضوضاء. يتالف النظام المقترح من اربع مراحل رئيسية هي تحويل المويجة المتقطعة ،ازالة الضوضاء باستخدام المويجة ،مرحلة الانصهار ومعكوس تحويل المويجة المتقطعة. في البداية تم تطبيق تحويل المويجة المتقطعة على مجموعة من الصور بنفس الحجم والمنظر مع استخدام المستوى الثاني من تحليل المويجة ومن ثم تطبيق احد تقنيات ازالة الضوضاء باستخدام التعتيبب على اجزاء معاملات المويجة وليس على الصورة باكملها من خلال استخدام قيمة عتبة لمعامل التفصيل ومقارنتها مع معاملات المويجة حيث تم استخدام العتبة الناعمة والصلبة مع عتبات الشامله وبايز حيث تتم مقارنة كل جزء من اجزاء الصورة مع الاجزاء المقابلة لصورة اخرى لاختيار الجزء الذي يحتوي على اقل ضوضاء. بعد ذلك يتم اختيار الاجزاء المثالية التي تحتوي على اعلى قيمة نسبة الاشارة الى الضوضاء من صور مختلفة لتكوين الصورة المنصهرة . والمرحلة الاخيرة هو تطبيق معكوس تحويل المويجة المتقطعة على الصورة المنصهرة لتحويلها من المجال الترددي الى المجال المكاني للحصول على الصورة الناتجة.الصورة الناتجة هي الصورة الجديدة التي تكون اكثر ملاءمة لرؤية الانسان والادراك الالي واظهرت نتائج هذا التطبيق ان طريقة العتبة الناعمة لازالة الضوضاء تعطي نتائج افضل من العتبة الصعبة بناءا على المقياس المستخدم. يعتبر النظام المقترح غير مناسب لدمج صورتين لان نسبة الاشارة الى الضوضاء للصورة الاولى والثانية تكون متساوبة. ان تطبيق الصور الضبابية على نفس النظام المقترح لايعطي نتائج مرضية في المجال الترددي حول دمج الاجزاء المثالية للصورة ولهذا السبب تم تطبيق الصور الضبابية في المجال المكاني مع مرحلة الانصهار باستخدام التصفية العكسية والتي اعطت نتائج جيدة للصورة الناتجة . الصور التي تم استخدامها في الرسالة هي صور رمادية ذات ابعاد مختلفة مع استخدام انواع مختلفة من الضوضاء. تم تنفيذ النظام المقترح باستخدام لغة برمجة (فيجوال ستوديو2012 ) vb.net. | The noise is unwanted information in image that is corrupted during its transmission or acquisition. The noise effects reduce the effectiveness of vision. Removing noise from the image eases the processing. The problem of the noise removal in the spatial domain is smoothing the data and the edges. Wavelet denoising is the process noise removal using wavelet in frequency domain. Therefore, it uses in keeping on the edges of the image, preventing types of noises and keeping the significant features of image, but this problem causes a loss in some of important information of image and produces image less visually. In the present time, image fusion is considered as one of types of integrated technology information, it plays a significant role in several domains and production of high quality images. The goal of image fusion is blending information from several images, also it is fusing and keeping all the significant visual information that is existed in original images and restoration the image quality that is based on PSNR and SNR measures. The proposed system includes four main stages are a discrete wavelet transform (DWT), denoising wavelet, fusion stage and inverse discrete wavelet transform (IDWT). In the beginning, discreet wavelet transform applies on the set of images of the same scene and size with two decomposition level. Then is performing the denoising wavelet hard and soft thresholding on sub band of image and is not the whole it, it is achieved by threshold value (universal and bayes thresholds) for detail sub bands and compare it with wavelet coefficient. After that is selecting the optimal sub band that contain less noise based on PSNR and SNR values for different images to form the fused image. Finally, apply IDWT process to convert the fused image from frequency domain to spatial domain and is obtained the resulted image. The resulted image is a new image which is more appropriate for human vision and machine perception .The results of the application show soft thresholding method is better than of hard thresholding that is based on higher PSNR value and image visually. The proposed system is unsuccessful for fusion two images, because PSNR value of one image and another image is same. The blur images applying on same proposed don't give satisfying results in the frequency domain about fused the optimal sub bands, therefore; it is applied in the spatial domain with fusion process by using inverse filtering and getting on the good results. The images are utilized in this thesis are gray images that different dimensions with using different types of noise as (salt &pepper, gaussian, Poisson and speckle) and the proposed system is performed using visual basic.net 2012 programing language.

تطوير ادوات جديدة في نظام مودل == Developing New Tools in Moodle System

Author name: نهلة فليح حساني الشويلي
Supervisor name: كاظم حسن كبان الابراهيمي
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Dhi Qar
First pages:
Abstract: E - learning is the use of IT in learning and teaching. This usage improved the quality of learning and teaching through facilitating access to resources and remote collaboration among teachers and students. E - Learning makes it easier to create and distribute learning materials, independently of geographical or time limits. It allows students to exchange information and participation in activities.There are many systems (software) development for management of E - learning (LMS). These systems provide tools for simulation the traditional learning, development and support the electronic learning.LMS provides many tools to create, manage the examination, evaluate the grades and spread question types.The general aim of this thesis is to develop the LMS that has been increasingly used having widespread in E - Learning in many countries. This system is "Moodle" which designed to supporting electronic learning.Moodle system is not satisfying conditions of the traditional exam copybook that is used in examination committees in Iraqi universities in terms of security and accuracy."Single view" report in Moodle system allows the teacher to edit grades of "Quiz" activities.Moodle system not provided question type in which the answering method is done through selecting word from a given text; this idea is a suggested product for the development of Moodle website in 2014 and doesn't have solution yet.In this thesis, we proposed three main proposals : 1. Development of Moodle system to use it in the examinations in Iraqi Universities through designing and implementation of new module named "Teacher view" to provide "The Electronic Exam Copybook" which simulated the traditional exam copybook that is used in the examination committees in these universities for assessment of student's answers. It also presents documentation of detailed information about student's attempt on examination and calculates the student's grades. This "Electronic Exam Copybook" satisfies all conditions of the traditional exam copybook in terms of security and accuracy. Also, the proposed module provides students' results in English and Arabic languages and manual of this module.2. Overcome of the weakness in "Single view" report in Moodle system.3. Development of Moodle system through designing and implementation of a new question type called "Highlight words" in this system. The idea of answering method to this question type is highlight word from a given clickable text.We used the following requirements to design and implement the proposals tools in this thesis : 32 bit windows7 operating system, 2.30 GHz Core i3 processor, router device, WAMP server, Moodle system (version 2.9) in English language, Mozilla Firefox web browser, Macromedia Dreamweaver 8 and First object XML Editor software

تصنيف اورام الدماغ باستخدام الشبكات العصبية الاصطناعية

Author name: نور حيدر داخل
Supervisor name: حازم باقر طاهر العلي
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Dhi Qar
First pages:
Abstract: Accelerating the pace of scientific progress and that focus in the field of human service requirements is to provide daily health and service. Health aspect is one of the most important things of science that needs to be develop it because it has an important effect on human life. Medical imaging has become one of the most important factors in the treatment of human health. A large amount of information and data can be obtained about disease by different methods of Images acquisition techniques same like Magnetic Resonance Imaging (MRI), Computed Tomography (CT).The purpose of this thesis is to discover a way to detect the brain is to suggest a system to diagnose the brain whether it is normal or abnormal (contains a tumor) by using artificial neural network technology which contain many steps. The first step of detecting the brain tumor system is the input image to the system as digital data.The processing is the second step, which includes reducing noise and removing impurities of the brain image. So as the noise is removed by using median filter and wiener filter. After processing (filtration) image segmentation. Feature extraction is the techniques that is used to measure of difference properties of image segments, given the specific features as input to the artificial neural network classified (ANN).Seeded classified databases are due to the two types of brain image, (normal or abnormal).Samples have taken from a group of patients in AL - Husain Teaching Hospital disparate ages and both ganders, which was 100 samples ,80 of them were training samples and 20 were tested .The proposed system has proven successful of the diagnostic through the results obtained on the ratio of (100% in the training phase) and (100% in the testing phase

تسريع ضغط الصور الكسوري باستخدام تقنيات مطورة == Fast Fractal Image Compression using Adaptive Techniques

Author name: حيدر عباس محسن
Supervisor name: حازم باقر طاهر العلي
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Dhi Qar
First pages:
Abstract: Fractal Image Compression (FIC) is one of the lossy techniques. In order for fractal compression image to be encoded is partitioned into non - overlapping blocks called ranges. From this range pool creates a new array called domain pool, the data of domain is array produced from taking an average of every four (2x2) adjacent elements in the range array, the domain is divided into overlapping blocks. Each block in range pool should be matched with every large number of blocks in the domain pool in order to find minimum error and recorded its best IFS approximate. The decoder process applies the determined IFS transformations on any initial image, and the process is repeated many times until reaching the attractor. In this thesis, it presents reducing the long encoding time of FIC without making significant sacrifice in the quality. In addition to the traditional fractal compression method, four IFS coding methods have been tested.The first method is speed up encoding phase by using moment features of the domain blocks with threshold , this method it compute the moment features of domain blocks only , during the matching process, compares the ratio factor of domain block with threshold, then compute the scaling (s), offset (o) coefficients and minimum error between range and domain blocks, and storing the optimal IFS coding , the results of this method with lenna image (4x4) block size, (the encoding time = 50.29 sec), (compression ration= 4.57 ) ,(PSNR=24.04), (RTR=33.42 ). The second method is speed up encoding phase by using the entropy technique of the domain blocks only , during the matching process, compares entropy value of domain block with threshold, and storing optimal IFS coding , the results of lenna image(4x4) block size, (the encoding time = 95.47sec), (compression ration= 4.47 ) ,(PSNR= 24.08) and (RTR= 17.60) . In the third method it merges the entropy technique with moment features of domain blocks only , during the matching process, compares entropy value and moments ratio factor of domain blocks with threshold and storing IFC coding, the compression results of lenna image (4x4)block size , (encoding time = 100.90 sec), (compression ration=4.47) ,(PSNR= 24.11) and (RTR=16.66) , the flaw in this method is the encoding time of 16x16 partitioning more than 4x4 and 8x8 partitioning. In the fourth method speed up encoding phase by use moment features of the range and domain blocks, in this method it computes the moment features of range and domain blocks, during the matching process, if the moments ratio factor of range and domain blocks are equal then compute scale(s),offset(o) and error (R) .storing best IFS coding. the results of lenna image(4x4) (encoding time = 114.02 sec), (compression ration= 4.47) ,(PSNR= 24.43), (RTR=14.74), the flaw in this method is the encoding time of 16x16 partitioning more than 4x4 and 8x8 partitioning. Matlab R2008a was used as a program to designed and implemented to achieve the coding and testing tasks. The environment used for our tests is a single PC with processor (core i3 CPU 2.53GHz) and 4.00 GB RAM .

خوارزمية موازنة الاحمال المقترحة في الحوسبة السحابية == A Proposed Load Balancing Algorithm in Cloud Computing

Author name: اطياف ضاري برغش
Supervisor name: خلدون ابراهيم عارف
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Dhi Qar
First pages:
Abstract: Cloud computing considers an important paradigm that provides services and dynamic virtual resources to the consumers over the internet.One of the main challenges and important fields for research in the cloud computing environment is load balancing. So the main goal of this thesis is to establish an effective load balancing algorithm for task scheduling. In cloud computing environment, the demands are executed into the virtual machines (VMs) and in the execution, it must be avoided a situation some of the resources are under loaded states while there are resources overloaded or heavy while achieving less the execution time of the tasks andresponse time. In this thesis, we present the Load Balancing Decision Algorithm (LBDA) which aims to achieve load balancing for reducing the : Total completion time (Makespan), Response time and Total execution time of tasks in cloud computing environment.The mechanism of LBDA is based on three functions, first : calculating the virtual machine capacity, virtual machine load then determining VM states (Under loaded, Balance, High balance, Overloaded) based on comparing VM capacity, VM load and using three thresholds : threshold upper; threshold fair; and threshold lower. Second : estimating the time required to complete executing the task in each VM. Finally, making a decision to distribute the tasks among the VMs based on VM state and task time required. Where checked task time required into VMs under states and selecting the fastest VM, when there are more than one VMs have the same time to complete a task, LBDA assigns a task to VMs which has the largest capacity. When all VMs become (Balance or High balance) states, the tasks are distributed between VMs depending on the fastest execution. The proposed LBDA was compared with Max - Min, Shortest Job First and Round Robin algorithms. We applied many experiments with various VMs’ capacity and : A different number of tasks on a different number of VMs, The same number of tasks on a different number of VMs, and A different number of tasks on a different number of VMs.The performance evaluation results implemented in cloudsim toolkit 3.0.3 and the experiments demonstrate that our proposed LBDA outperformed to other algorithms by reducing the completion time of tasks, response time, andtotal execution time of tasks.

مخطط يربط المفتاح بالقياسات الحيوية لحماية انظمة المعلومات == A Biometric Key Binding Scheme for Information Systems Protection

Author name: ابرار خلف الشريفي
Supervisor name: كاظم مهدي الموسوي | كاظم حسن الابراهيمي
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Dhi Qar
First pages:
Abstract: Cryptosystems becomes an increasingly important feature as the most reliable tool in network and information security. However, while there are many forms of cryptosystems exist today a weak link of all systems is the secure management of the cryptographic key. And as a result of the growing interests in biometrics a new field of research has emerged, which is called "biometric cryptosystems". Within biometric cryptosystems the advantages of biometric authentication are introduced to generic cryptographic key management systems to enhance security.This work proposes a scheme to bind a cryptographic key with the face biometric by combining correlation filters based biometric recognition with a biometric key binding scheme with providing the variation tolerance, discrimination and security which are the most important requirements that apply to a cryptographic key retrieval using a biometric. In the proposed scheme the cryptographic key is generated independently of user's face image using Pseudo - Noise generator to produce a unique and long enough key (128 bits) for each user then this key is linked to the face image only through a secure block of data known as the 'protected record'. The correct key will only be derived via the interaction of this protected record with the correct user's face image during a live authentication process. The resultant key may be used in a system as cryptographic key or as a personal identification number (PIN) to overcome the need to carry, store, or remember keys for cryptosystems or any other application that are used (PIN).The proposed scheme performance has evaluated using FACE 94 facial database. The performance rate reached to 100% which is a perfect result for practical applications. The proposed scheme is implemented using (MATLAB 2008). It is executed for testing purpose on computer with processor of 1.80 GHZ dual core, i3 under Microsoft Windows 7 operating system.

اداة تقييم الخطر لتنبؤ احتمالية الاصابة بسرطان الثدي بالاستناد على عوامل الخطر الديموغرافية واستخدام خوارزمية الشبكة العصبية الضبابية == Risk Assessment Tool - based Demographic Risk Factors to Predict Probability of Breast Cancer accident using Neuro - fuzzy Technique

Author name: دنيا سمير مزهر
Supervisor name: كريم راضي حسن
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: Arabic
University location: Basrah
First pages:
Abstract: In recent years, the statistics presented high incidence rates of breast cancer.So, early detection of breast cancer considered one of the highest healing rates for most cases, where the treatment at early stage is very easy and involves removal to malignant tumor from breast without needed to remove all breast. As well the cost of treatment is very economical with early diagnosis, while at a late stage, the cost associated with treatment is higher. In addition, the prevention have role to reduce risk by avoid catalysts that cause Incidence and associated with risk factors. For these reasons and others, we present the proposed method for designing a risk assessment tool for breast cancer prevention and early detection based on demographic risk factors (DRF) and using Neuro - Fuzzy System (NFS) in designs risk assessment model that help doctors and patients in assess risk of Incidence through providing better and more reliability results.In this thesis, the proposed method is a process of several stages. The first phase included statistical study to identify problem of research based on statistics in Basra, through apply the study on patients community at Al - Sadir Teaching Hospital in Basra (Oncology Center), that represent main center in Basra for the treatment of cancer. Three and most important phase includes the proposed approach for design assessment model using Neuro - Fuzzy system (NFS) in two approaches. the second approach suggested using Fuzzy c - means (FCM) algorithm with Neuro - Fuzzy system to improve the first approach that just use Neuro - Fuzzy system (NF). The proposed method is focused to solving basic research problem to prevention and early detection of breast cancer.The proposed method achieved high results when compared to other methods and tools. The success rate of proposed assessment tool is 94%, and it is higher rate among other assessment tools (Gail, IBIS) that obtained by match the result for assessment of test data (50) with the expert opinion (the real result). Programming and testing of proposed method by used Matlab 2013

اخفاء معلومات في ملف صوتي باستخدام تقنيات التشفير الانسيابي والمواقع العشوائية

Author name: اسماء عزيز جابر
Supervisor name: حميد عبد الكريم يونس
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: Arabic
University location: Basrah
First pages:
Abstract: The rapid development and tremendous computer and communication technologies and the proliferation of global networks, has led to the emergence of many techniques to transfer confidential information including encryption (Cryptography), and as a result of piracy And penetrating these techniques need arose for techniques and methods more sophisticated and difficult to transfer confidential information securely. Because of the increase in the number of attacks recorded during the exchange of electronic information between the source and the recipient may need to use a more powerful way to secure data transfer, encryption methods are combined And hide the information together for the purpose of setting up a mechanism to improve confidentiality and because these methods shared goals and common services in protecting the confidentiality, integrity and availability of information from unauthorized access and thus check mechanism High security integration to preserve data confidentiality.The research aims to use the method to hide secret information (picture or sound) into audio files (audio file) (wav) as cover in such a way as to ensure that information is not piracy, secret message encryption Sleek encryption before concealment by making the size of the stego - audio is equal to the size of the original file.As the audio file is the output of a process of hiddin similar to the original file and for the human ear to feel the change. By segmenting the original soundtrack cover frames (frames) and apply the random function to locate hidden after the colored image layers in case confidential information picture or sound to hide the hash to Frameworks and then applying the function for selecting locations for concealment and then apply the least significant bit technology (Siginfcant Least Bit) for concealment.In the case of the recovery process is contrary, embedded audio file fragmentation occurs (stego - audio) and apply the random function for selecting locations for concealment and decrypt the message.Use a key in a hash process sound represents the size of the picture you want to hide, and it represents the number of frames (frames) and serve as a key to encrypt the original audio as we use an algorithm to generate a sequence of key Homologous between the sender and recipient to Sleek encryption to increase security. The recovery process success ratio for the image dB 99.0, audio retrieval process was well proportioned. , We used Image metrics (PSNR), (SNR) and the inclusion ratio between the confidential letter and sound and cover (RMSE) between audio - stego and original sound. And high results were obtained about (85 - 95 )when recovery without any mention on the file (Stego_audio).

تصميم وتطبيق نظام قواعد بيانات موزعة لادارة المعلومات الصحية وبناء نظام لاستيراد البيانات == Design and implementation of distributed databases system for management health information And build a system to import data

Author name: محمد قاسم ياسين
Supervisor name: طالب عبد الصمد عبيد
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: Arabic
University location: Basrah
First pages:
Abstract: استدعت زيادة المعلومات في المجالات الحياتية كافة الحاجة الى استخدام الادارة الالكترونية وتحويل جميع المعلومات الورقية الى معلومات محوسبة تمتاز بسهولة استرجاعها والاستفادة منها في عمليات اتخاذ القرار. نحتاج للاستفادة من المعلومات المحوسبة في اي مكان وزمان الى طريقة تمكننا من ترتيبها وفهرستها. ظهرت مجموعة من التقنيات لخزن المعلومات المحوسبة واسترجاع منها نظم الملفات ونتيجة التطور الحاصل في معدات الحاسوب والبرمجيات ظهرت انظمة ادارة قواعد البيانات المركزية، وانظمة قواعد البيانات الموزعة واحتلت الاخيرة اغلب تطبيقات قواعد البيانات لما تقدمه من تقنيات مختلفة في تخصيص البيانات والاستفاده منها في اماكن مختلفة. تعد الصحة الالكترونية احد انماط الادارة الالكترونية والسجلات الصحية الالكترونية احدى اشكال الصحة الالكترونية التي تعتمد عليها عملية تقديم الرعاية الصحية داخل القطاع الصحي. تم في النظام المقترح تصميم وتطبيق قواعد بيانات موزعة لغرض ادارة معلومات المراجعين المرضى خلال زياراتهم للدوائر الصحية (المراكز الصحية والمستشفيات)، من خلال استخدام تقنيات التجزئة والتكرار في توزيع البيانات بما يحقق استقلالية وسيطرة للوصول للبيانات.احتوى النظام قاعدة بيانات للمراكز الصحية (وتجزئتها وتكرارها حسب قطاعات المناطق) وقاعدة بيانات للمستشفيات يرتبطان معا لتكامل البيانات وجعل كافة المعلومات امام المستخدم.يكون الوصول الى البيانات من خلال الشبكة المحلية او من خلال الانترنيت باستخدام نظام اسماء النطاقات الديناميكية (DDNS) Dynamic Domains Name System او الرقم العام Public - IP.بسبب الحاجة الى الاستفادة من البيانات المتواجدة في الانظمة القديمة والتي تختلف في نظام الادارة او هيكل البيانات, تم بناء برنامج تطبيقي لنقل البيانات بين قواعد البيانات المتماثلة وغير المتماثلة على حد سواء. يحتوي البرنامج التطبيقي على مجموعة من اجراءات المعالجة تستخدم لنقل البيانات من قاعدة بيانات الى اخرى قد تختلف في نظام ادارة قاعدة البيانات او هيكل الجداول او الانواع البيانية للحقول وامتاز نظام ادارة معلومات المرضى على الامكانيات الاتية : - السماح بعمليات الاضافة والحذف والتعديل والبحث. - متابعة اماكن تواجد المريض من لحظة دخوله الى المستشفى حتى خروجه. - امكانية الوصول الى البيانات من خلال الشبكة المحلية او من خلال الانترنيت. - امكانية تصدير معلومات مراجعة المريض الى ملف اكسل. - متابعة انتشار الامراض خلال فترات من الزمن من خلال مخططات البيانات Data Charts. - استخدام بريد الكتروني داخلي بين المستخدمين. - استيراد البيانات من قواعد بيانات مختلفة بنظام ادارة قاعدة البيانات او الهيكل البياني والتنظيمي للجداول. - احتوائه على قسم خاص لاضافة الصلاحيات للمستخدمين بالاعتماد على الوظائف المخوله لهم. - تنبيه مدير النظام ان هناك محاولات للوصول الى النظام برمز مرور خاطئ

عنقدة النصوص العربية باستعمال تحليل الاحالة المشتركة == Arabic Text Clustering Using Coreferance Resolution

Author name: فراس حمودي نعمة
Supervisor name: سلمى عبد الباقي محمود
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: Arabic
University location: Basrah
First pages:
Abstract: Clustering texts organizes texts in subsets coherent and internally consistent, and different with each other. These subsets called clusters. The documents combines according to the similarity measures, that rely on features extracted of the documents. This technique applies in various fields such as web mining, search engines, and information retrieval. Clustering documents gives information retrieval, automatic extraction, and representation efficiently without user intervention for increasing growth of the new documents.This study aims to improve the accuracy and efficiency of arabic texts clustering. To achieve the aim of this study two approaches are used. The first approach uses the standard technique in Arabic texts clustering. The second approach is proposal to reduce features of the texts.The first approach is applied (K - medoids, K - means) and similarity measures (Euclidean, cosine). The problem of this approach is that huge of the features, which influence the efficiency and coherence of the clusters.The huge features for the documents adds challenge and lead to high dimension so coreference resolution technique approach is applied. This technique extracts main subjects for each document to improve arabic documents clustering in order to achieve the goal of our study. The system implements using a corpus contains on 200 sport news Arabic.Finally, evaluation measures are used including (Precision, Recall and F - measure) to evaluate our system, and we obtain acceptable results.First method exceed K - medoids and using similarity measures (Euclidean, Cosine) where highest values (0.60, 0.78, 0.67) of the method of k - means. In addition, the results of the proposed second way exceeds method of K - medoids and using similarity measures (Euclidean, Cosine) with coreference resolution, where the measures values (0.80, 0.83, 0.81).

تصميم خارطة مفاهيم لبعض مفردات اللغة العربية == DESIGN AN ONTOLOGY FOR SOME ARABIC VOCABULARY

Author name: حنين عقيل لفتة
Supervisor name: عبد الحسين محسن عبد الله
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: Arabic
University location: Basrah
First pages:
Abstract: Since the growth and development of technology, there have been many changes to visualize the current Web. After decades of Web's using , there were demands raised to restructure the web in order to allow machines do the work of the man on the web. So, the aim of this main generation idea is to enable machines .These machines could understand what they display by adding meaning to the traditional Web content, this is called the Semantic Web.The form of this meaning seems in a huge global database that has information connected to each other and understand by the machines. Therefore, the machines can be realized the relations between information, analysis and indexing of knowledge categories. All these which make the machines are responsible on main part of information research process, while the human's part is just to receive the results that are almost ready to benefit from them. Ontology notion seems to express a set of concepts within a particular field as well as the relationships between these concepts.In spite of the appearance of the Semantic Web, ontologies spread widely in various languages, but there was a shortage in Arabic language ontology within different areas. Even when it found, it was small attempts and unavailable on the linked data form in the web.This work aims to design an ontology that cares about the Arabic language and its components. So that this project focuses on these things : 1. Create a model for an Arabic Language Ontology, as follows : a. Identify the main concepts for the Arabic language like (noun, verb, letter, adjective and adverb) and their properties to form the structure of the ontology.b. Modeling the resulting ontology in one of modeling language .It has been used Object Role Modeling (ORM) because of its high strength to express and transform this model to the language The Web Ontology Language (OWL) .c. Verifying the ontology's components are logically connected and provide the right answers to the questions that are designed to by asking queries to it in SPARQL language.2. Populate ontology with individuals by - : a. Collect several instances from different sources and clarify their types by linking them to the main concepts of the Arabic language.b. Extract every meaning of the instances and linking them to the same meaning in the English language for the clarification .This is to indicate the possibility of using this ontology as dictionary bi - languages.That is one of its applications by linking it to other ontologies to reuse their information and exchange them.3. Sharing this ontology on WWW using Linked Data technology with free licenses.4. Creating a web page that contains Arabic - Arabic language dictionary in HTML and JavaScript languages .These were achieved by using the same words that included in the proposed ontology .All these give their meanings and types and be availableto the average user .5. The conclusion of this ontology is specifying many meanings to one word, however the ontology give us one type for specific meaning. That is considered negatives of the proposed work.Keywords : Semantic Web, Ontology, Ontology Population , Linked Data, ORM, OWL, SPARQL

Secure Email Message Using Genetic Algorithm - Based Elliptic Curve Cryptography Systems

Author name: محمد حسن حالوب
Supervisor name: Haider M. Abdul - Nabi
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Basrah
First pages:
Abstract: Email is a service for exchanged messages between group of people (sender and receiver) using communication network. Nowadays people, companies, and business sectors are using email in the official communications. There are a lot of threats on the email service and the most importantly one is the eavesdropping. Some email messages are very important and must have a level of security during exchange of them.The current study uses hash function to add data integrity by applying Secure Hash Algorithm - 1(SHA - 1). This study uses Elliptic Curve Digital Signature Algorithm to add authentication and non - repudiation to the email message sender.Encrypt of an email message is a solution to overcome the eavesdropping. Hence, the attacker can obtain the email message but cannot read it without the key. The encryption algorithm, which is used, is Elliptic Curve Cryptography. This algorithm has a strong encryption, high level of security, and small key size compared to the RSA algorithm.The contributions of this thesis are : proposes a new method based on genetic algorithm to select the most effective parameters of the Elliptic Curve Cryptography (a,b); and proposes four encryption methods : updated symmetric Elliptic Curve Cryptography, modified Menzes - Vanston Elliptic Curve Cryptography, hybrid symmetric and asymmetric Elliptic Curve Cryptography, and hybrid Menzes - Vanston Elliptic Curve Cryptography and AlGamal Elliptic Curve Cryptography.The proposed methods are compared against previous techniques. It is found that these methods outperform the previous techniques in terms of the encryption/decryption processing time and the level of security.

اخفاء المعلومات في نص عربي باستخدام تقنيات معالجة اللغات الطبيعية == Information Hiding In Arabic Text Using Natural Language Processing Techniques

Author name: عبد الرحيم عبد الصاحب عبد الرحيم
Supervisor name: سهاد مال الله
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Baghdad
First pages:
Abstract: التطور السريع للاتصالات جعل من الصعب للحفاظ على المعلومات في مامن من الاشخاص غير المخولين. بعض المعلومات مهمة وتحتاج الى ان تصنف على انها معلومات سرية مثل الاستخبارات، الاعمال التجارية، والعسكرية... الخ، وتستهدف هذه المعلومات من قبل منافسي المرسل او المت | The fast evolution of communications makes it hard to keep information safe from the unauthorized person. Some information is important and needs to be classified as secret information such as information about Intelligence, Business, Military...etc., th

تشفير الصوت الحصصي باستخدام تحويل الجيب تمام المنقطع == Share Audio Cryptography Using DCT Transform

Author name: سارة سعد علي البغدادي
Supervisor name: عبير متي يوسف
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Baghdad
First pages:
Abstract: الحماية الفعالة والامنة للمعلومات الحساسة هو مصدر القلق الاساسي في نظم الاتصالات او انظمة التخزين الشبكي، فمن المهم لاي معالجه معلومات هو ضمان عدم العبث بالبيانات.و لتحقيق السرية وسلامة المعلومات والوسائط المتعددة، تم وضع عدة مخططات الحصص السرية المختلفه | An Effective and secure protection of sensitive information is the primary concern in communication systems or network storage systems, it is important for any information process to ensure data is not being tampered. To achieve confidentiality and integr

تصميم وتنفيذ روبوت مخصص لانجاز مهام خاصة بالهندسة المدنية (عملية طلاء الجدران) == Design And Implementation Specified Robot For Civil Engineering Work (Wall Painting Process)

Author name: يونس صابر عثمان
Supervisor name: عبد الستار محمد خضر | معن محمد شاكر
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Mosul
First pages:
Abstract: يعتبر تصميم وبناء روبوت تحدي كبير ويتطلب الجمع بين علوم مختلفة مثل الالكترونيات والسيطرة والبرمجة والميكانيك والاتصالات والفيزياء. كما ان تجميع وفحص اجزاء الروبوت يتطلب جهد كبير.الخطوة الاولى في هذا البحث هي تصميم روبوت متنقل يقوم بعملية طلاء جدران قاعة. | The design and implementation of a robotic system is a very big challenge. It requires different integrated fields of sciences, like electrical, control, programming, mechanical, communications, physics, and others. The assembling and testing of the parts

تشفير الرسائل القصيره باستخدام RC4 مع مفتاح متغير بالاعتماد على الاندرويد == Sms Encryption Using RC4 With Dynamic Key Based on Android

Author name: سجى طه احمد عبد
Supervisor name: لؤي ادور جورج
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
University: University of Baghdad
Language: English
University location: Baghdad
First pages:
Abstract: تعد خدمة الرسائل القصيرة (اس ام اس) ھي جزء من خدمة الاتصالات المستنده على ارسال النصوص التابعة لانظمة الاتصالات المتنقلة. وبصرف النظر عن الدردشة الالكترونية، اصبحت خدمة الرسائل القصيرة اليوم مصدرا اساسيا لتبادل المعلومات السريھ او الخاصة، كما انھا تلعب دو | Short Message Service (SMS) is a text based communication service component of mobile communication systems. Apart from electronic chatting, SMS has become a fundamental source for communicating confidential or proprietary information. Also, it plays very

تقسيم وتمييز النص العربي المكتوب يدويا == Arabic Hand Written Segmentation And Recognition

Author name: منى خلف ضمد
Supervisor name: بان نديم ذنون يونس | عماد جاسم محمد
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
University: University of Baghdad
Language: English
University location: Baghdad
First pages:
Abstract: The ultimate aim of handwriting recognition is to make computers able to understand human written texts, with a performance comparable to that of humans. Arabic handwriting recognition is an active research area in pattern recognition and has many practic

تجنب الحواجز للروبوت المتحرك داخليا == Barrier Avoidance For Indoor Mobile Robot

Author name: ضحى عبد الله عامر
Supervisor name: محمد صاحب مهدي الطائي | لؤي كاظم عبود
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
University: University of Baghdad
Language: English
University location: Baghdad
First pages:
Abstract: The motivation of current research is addressed in establishing a robotic control system to follow a line was drawn on the ground, and then this system has been programmed to avoid the barriers on its way. Thus, the proposed robotic system is designed to

اقتراح نموذج التعدين لتحليل الشبكات الاجتماعية على الانترنت (الفيسبوك) == Proposed Mining Model For online Social Network Analysis (Facebook)

Author name: سراء موفق عبود
Supervisor name: سكينة حسن هاشم
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Baghdad
First pages:
Abstract: شبكات التواصل الاجتماعي على الانترنت, في بعض الاحيان تسمى وسائل اجتماعية, اصبحت جزء اساسي في حياة معظم الناس اليومية من عملية فحص حسابات الفيسبوك والتويتر حتى عملية النشر فيها, وقائمة الاهتمامات وتحميل اشرطة فيديو اليوتيوب يمكن لهذه الشبكات الاستفادة للمش | Online Social Networks (OSN), sometimes called social media, have become an essential part of most people’s everyday lives, from checking Facebook and Twitter to posting blogs, interesting listings, and uploading of YouTube videos. These networks can leve

نهج جديد للمحادثة الامنه في الوقت الحقيقي == New Approach For Security Chatting In Real Time

Author name: سرى خلف اخباله
Supervisor name: علاء كاظم فرحان
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Baghdad
First pages:
Abstract: ان اي منظمه لها موظفين فانها تكون بحاجة الى نظام اتصالات فيما بينهم ويمكن استخدام نظام محادثة خاص يعمل على الشبكة الداخلية للمنظمة )انترانت( او عام ويعمل على شبكة الويب )الانترنت( لتبادل المعلومات والاستعلام فيما بينهم. فالمحادثة هي المشاركة المتزامنه في | Any organization having any number of employees needs a communication system which could be used for private conversation on the organization's internal network (intranet) or in working on the Web (Internet) to exchange information and inquiry among its m

تصنيف صور الماموكرام باستخدام التحويلات الكنتورية == Mammogram Images Classification Using Contourlet Transform

Author name: زينب فوزي عبد
Supervisor name: مثيل عماد الدين عبد المنعم
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Baghdad
Key words:
  • Mammogram Image
  • Classification
  • Contourlet Transform
  • Wavelet Transform
  • Otsu's Thresholding
First pages:
Abstract: Breast cancer is the most common malignancy in adult females and is the second most common disease that may lead to death. Early detection is important for manage breast cancer that give a better chance of full recovery. The detection rate can be increase

خوارزمية تصنيف شجرة القرار الامثل لمجموعة بيانات == An Optimized Decision Tree Classification Algorithm For A Data Set

Author name: احسان ا. كريم
Supervisor name: Mehdi G. Duaimi
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
University: University of Baghdad
Language: English
University location: Baghdad
First pages:
Abstract: A decision tree is an important classification technique in data mining classification. Decision trees have proved to be valuable tools for the classi?cation, description, and generalization of data. Work on building decision trees for data sets exists in

ضغط البيانات باستخدام تمثيل الترميز التسلسلي == Data Compression By Using Chain Code Representation

Author name: فنر علي جودة
Supervisor name: توفيق عبد الخالق الاسدي
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Babylon
First pages:
Abstract: In this thesis we have proposed an efficient lossless image compression system. This system consists of two parts : the first part is compression procedure that is based on two approaches of chain code representation. The first approach is Variable Vertex

طريقه جديدة لكشف واستخراج التزوير من اللوحات الفنية == New Method For Detect And Extract Forgery From Paintings

Author name: سبا راكان سالم
Supervisor name: هناء محسن احمد
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
Language: English
University location: Baghdad
First pages:
Abstract: نظرا لتوفير العديد من برامج تحرير الصور وادوات معالجتها, مما ادى الى سهوله التغير على المعلومات التي تحملها اللوحات دون ترك اي اثار واضحه عليها جراء العبث بها والتي ادت الي مشكله التحقق من الصور. ومما ادت هذة القضايا المتعلقة بامن الوسائط المتعدده في تطو | Due to availability of many image editing and processing tools, it is possible to easily change the information represented by a digital paintings without leaving any obvious traces of tampering, which led to the problem of verification the image.These i

مقياس اسلوب المؤلفات الادبية للنصوص العربية == Stylometric Authorship Arabic Text Detection

Author name: عمار عادل عبد الرزاق
Supervisor name: طريف كامل مصطفى
General topic: Computer Science
Specific topic: Computer Science
Degree: Master
University: University of Baghdad
Language: English
University location: Baghdad
First pages:
Abstract: ان التدوين والنشر الالكتروني جعل عمليات السرقة متاحة وسهلة، وخاصة في مجال المؤلفات الادبية، لذلك اصبح من الضروري وضع بصمة خاصة للمؤلف لمنع عملية الانتحال. التحليل والتنقيب عن النص يعتبر من اهم السمات لقياس اسلوب المؤلف، على سبيل المثال، المؤلفات والرواي | The electronic publications made the theft operations easy and accessible, especially in the field of the literature; therefore, it became necessary to put a style print for the author to prevent the plagiarism.Stylometric authorship attribution is conce
1 ... 61 62 63 64 65 ... 80