Yüksek Lisans Tezleri

Permanent URI for this collectionhttps://hdl.handle.net/20.500.12416/15956

Browse

Filters

2003 - 200922010 - 201915

Settings

Access Right: Open AccessDepartment: search.filters.department.Çankaya Üniversitesi,Fen Bilimleri Enstitüsü, Bilgisayar Mühendisliği Ana Bilim Dalı

Search Results

Now showing 1 - 10 of 17
  • Master Thesis
    Constructing low diameter topic-based Pub/Sub overlay network with minimum maximum node degree
    (2016) Layazalı, Sİna
    Publish/subscribe communication systems, where a large number of nodes (publishers and subscribers) are interested in variety of topics, have received more attention in recent years. Constructing an efficient overlay network that connects the nodes interested in several topics is an important issue in these systems. In constructing effective and scalable overlay networks, pub/sub network designers prefer to keep the diameter and maximum degree of the network low. This is in addition to lower running time cost and higher scalability. However, the existing algorithms seem to fail in decreasing maximum node degree and reducing diameter. To address these issues for a given set of nodes with variety of topics, we present a heuristic algorithm known as Constant Diameter Minimum Maximum Degree (CD-MAX) which decreases the maximum node degree and maintains the diameter of the overlay at most at two. The new algorithm improves the maximum node degree, by 64 percent. The CD-MAX algorithm selects the node with the lowest node degree and connects it to its neighbors. For more than two nodes with an equal highest node degree, this algorithm selects the node with the highest density. The CD-MAX algorithm has a refinement version that decreases the maximum node degree even further. The algorithm is validated and analyzed through simulations.
  • Master Thesis
    An exploratory study to assess analytical and logical thinking skills of the software practitioners using a gamification perspective
    (2015) Kayalı, Şahin
    The link between analytical and logical thinking skills and success of software practitioners attracted an increasing attention in the last decade. Several studies report that the ability to think logically is a requirement for improving software development skills, which exhibits a strong reasoning. Additionally, analytical thinking is a vital part of software development for example while dividing a task into elemental parts with respect to basic rules and principles. Using the basic essence of gamification, this study proposes a mobile testing platform for assessing analytical and logical thinking skills of software practitioners as well as computer engineering students. The assessment questions were taken from the literature and transformed into a gamified tool based on the software requirements. A focus group study was conducted to capture the requirements. Using the Delphi method, these requirements were discussed by a group of experts to reach a multidisciplinary understanding where a level of moderate agreement has been achieved. In light of these, an assessment tool was developed, which was tested on both software practitioners from the industry and senior computer engineering students. Preliminary results suggest that individuals who exhibit skills in analytical and logical thinking are also more inclined to be successful in software development.
  • Master Thesis
    Structure from motion using a single camera
    (2016) Isawi, Muthana Yaseen Nawaf.
    This thesis introduces a general survey of conversion algorithms, their advantages and disadvantages and a thorough explanation of the basic concepts in the field of 3D model reconstruction. The thesis concentrates, step by step, on the structures of motion technique and reconstruction of three-dimensional models from image pairs. The reconstruction process is carried out using a single calibrated camera and an algorithm based on only two views of a scene, the SFM technique based on detecting the correspondence points between the two images, and the epipolar inliers. All the experimental results have been computed using MATLAB (R2015b). By using the KLT algorithm we figure out the incompatibility of it with the widely-spaced images. Also, the ability of reducing the rate of reprojection error by removing the images that have the biggest rate of error. The experimental results are consisting from three stages. The first stage is done by using a scene with soft surfaces, the performance of the algorithm shows some deficiencies with the soft surfaces which are have few details. The second stage is done by using different scene with objects which have more details and rough surfaces, the algorithm results become more accurate than the first scene. The third stage is done by using the first scene of the first stage but after adding more details for surface of the ball in order to motivate the algorithm to detect more points, the results become more accurate than the results of the first stage. The experiments are showing the performance of the algorithm with different scenes and demonstrate the way of improving the algorithm. At last, we would like to mention that the aim of thesis is to figure out the depth information from two 2D images, and not to create 3D image from two 2D images.
  • Master Thesis
    Obstacles to applying e management in Libya
    (2019) Hussin, Ramzi
    E-management can eliminate the obstacles faced by traditional management by the use of information and communication technologies (ICT). ICT is widely used by government, public and private institutions in order to facilitate the life of people. The current study uses a survey research method in order to obtain opinions of the participants on obstacles of applying e-management in Libya. An online survey was presented to 120 participants to get self-reported quantitative data. Descriptive statistics were used to evaluate the perceptions of the participants on the obstacles of e-management. The overall driver assessment shows that e-management project in Libya is possible to fail at current time. The mean of responses on the strategy assessment is 3.18, suggesting that there is not enough strategy to apply this assessment at current time. The overall management assessment showed that the mean value was 3.34, which is less than 5. This result was expected due to the lack of successful management in governmental institutions. The analysis of the design assessment showed that it was close to the failure. In spite of the overall competencies assessment, the existence of one factor seems to succeed where it has a mean value of 5.28. This factor is the education of information and communication technology by system managers, developers, operators and users. The overall assessment of this section has a mean value of 4.81. The overall technology assessment has a mean value of 3.61, which is less than 5 and indicate that the technology assessment would fail if applied at current time. The final assessment of our study is the other assessments which have a mean value less than 5. This assessment has a one statement in which mean value of the assessment is greater than 5. This statement is about perseverance by implementers where the mean value of responses is 5.02.
  • Master Thesis
    Authentication mechanism of Electronic Health Record (EHR) in the cloud
    (2015) Hurmuzlu, Mine
    Electronic health record (EHR) is a system that contains patients' complete medical related records from birth to death. These records consist of diagnoses, medications, laboratory tests, and results. The records are created by the patients and accessed by health providers such as doctors, nurses, pharmacies, etc. The advantages of using cloud computing has helped the health organizations to dispense the local servers and shift all their data to the cloud, which helped them save space, energy, and cost, as well as provide the benefit of accessing the data by patients and healthcare providers from anywhere at any time. Exchanging medical records in the cloud, however, has threatened the security and privacy of e-health systems where the authentication, access control, and integrity of the medical records are the main challenges in e-health clouds. In this thesis, we study the authentication mechanism of an EHR system and investigate an improved version using attribute based encryption ABE. Patients have full control over their medical records which are stored in a semi trusted servers. The system also works on access policies in case of emergency.
  • Master Thesis
    Development of a face recognition system for e-government in Iraq
    (2014) Hassan, Morad
    Innovations of technology build important effects on the world wide. Technological developments influence economy, social life, business and the way public services delivered. Public services are provided in associate degree electronic format to scale back time and value and to boost service quality. These innovations also cause a social transformation. Governments are required to adopt such changes and to be a related of "Information Society". This thesis proposes e-government model based on the citizen's participation and improvements in the delivery of governmental services by using a qualitative case study strategy considering the Kirkuk region in Iraq. Regarding the proposal a Graphical User Interface (GUI) has been developed in order to compare the input face image with existing database images to display the citizen information such as (name, surname, birth date, etc.) This e-government model can also be used to detect the cases of forgery
  • Master Thesis
    A comparison of NoSQL database systems: A study on MongoDB, apache HBase, and apache Cassandra
    (2016) Hammood, Ali Hussein
    Due to their many useful features, database management systems have been used widely with relational data for over 20 years. However, such systems are not able to handle massive and complex data efficiently. New systems known as NoSQL database management systems have appeared to deal with massive and complex data that provide fast and high performance. In this thesis, we discussed and tested three kinds of NoSQL database system in order to reveal their capabilities and how they respond in different operations. For this purpose, we set up a novel testing environment for each workload and examine the responses for the three systems. The results of this study show the weaknesses and strengths of each database system used in the study. Due to the different architectures of each database that we tested, we have seen different responses for each with changed workload operations. In our work, we used the Yahoo Cloud Serving Benchmark (YCSB), which is a framework designed by Yahoo to test database performance. According to the results obtained, we can conclude that MongoDB performed very well with low throughput, but not as well with high throughput. Cassandra and HBase performed very well under heavy loads due to their optimized designs. In the read operation, HBase has poor performance as compared to other systems tested.
  • Master Thesis
    Multi-word expression detection for turkish
    (2016) Hürmeydan Ersoy, Nazlı
    In this thesis, I performed text analytics on Turkish academic articles about four science subjects and detected collocations according to statistical measures and tried to benchmark the results of each method in these subjects. The main purpose of my thesis is to create a terminology dictionary using only example journal articles. In accordance with this purpose, I applied some machine learning methods on Weka to test if my scores and Weka results can get more reliable results.
  • Master Thesis
    Sentiment analysis and gender prediction in twitter data
    (2015) Balaban, Ertuğrul
    In this thesis, tweets from Twitter that have been sent by users will be considered on a preferential basis in accordance with determined or requested specific key word(s). Also the interpretation of these tweets, by the computer, will be examined in a way as they are "Positive", "Negative" or "Neutral". In this context, under the heading 'Twitter Sentiment Analysis', studies were conducted and the success rates of achieved results were compared. In addition to this, on the basis of the usernames(of users) who send tweets, tweets was compared with Turkish Special Names which is shared by the Turkish Language Association (TDK) and also achieved results and gender determinations of users in terms of "Female", "Male" or "Not Determined," were examined. Under the heading of 'Gender Prediction in Twitter' studies were conducted and the success rates of achieved results were compared. On the basis of this study, the related topics of 'Sentiment Analysis' and 'Gender Prediction' were examined for Turkish Language and all of these studies were carried out through Turkish language.
  • Master Thesis
    Analysis of mammography images for cancer detection
    (2016) Alshana, Ghassan
    Mammography is the best available technique for early detection of breast cancer. The most common breast abnormalities that may indicate breast cancer are masses. Also, there are some signs that can lead to breast cancer diagnosis, such as architectural distortion and bilateral asymmetry. In this study, an algorithm is used to detect breast cancer in mammography images. Four stages are presented: (1) preprocessing, (2) segmentations of regions of interest (ROI), (3) feature selection and extraction, and (4) classification. In the preprocessing stage, the digital mammogram is pruned, 2D-median filter is used to filter the image and unnecessary labels are removed from the breast. In the segmentation stage, global thresholding is used for segmenting the breast. Morphological operations like erosion, dilation, opening and closing are used to enhance the breast. Seeded region growing is used for removing the pectoral muscle and for segmenting the mass in the breast. In the feature selection and extraction stage, intensity features are selected and extracted from the ROI. In the classification stage, the extracted features are fed into artificial neural network (ANN) classifier to classify the mass as malignant or benign. The output of the proposed method would assist radiologists to examine images containing unusual masses more closely and to help them minimize misinterpretation. The method achieved 91.30% sensitivity, 91.30% specificity and 91.30% accuracy resulting from the confusion matrix which is a performance evaluation metric.