Yazılım Mühendisliği Bölümü Koleksiyonu
Bu koleksiyon için kalıcı URI
Güncel Gönderiler
Öğe NCA-based hybrid convolutional neural network model for classification of cervical cancer on gauss-enhanced pap-smear images(Wiley-Blackwell, 2022) Bingöl, HarunCervical cancer is a very serious disease that deeply affects women's lives, often resulting in death. This type of cancer, which is very common in women, is diagnosed at an early stage and is of vital importance for the success of the treatment. Pap-smear tests are used by physicians as the primary diagnostic tool to diagnose the disease. In this study, a hybrid deep model is proposed to classify pap-smear images to detect cervical cancer. In addition, the Gaussian method was applied to improve the images in the original dataset. Feature maps were taken from both the original dataset and the Gaussian-enhanced dataset in the built hybrid architecture, which used Darknet53 and Mobilenetv2 models as the base. After these obtained feature maps were combined, useless features were extracted and the number of features was reduced by using the Neighborhood Component Analysis (NCA) dimension reduction method. Finally, this optimized feature map was classified into different classifiers. As a result of the experimental studies, it was determined that the proposed hybrid model performed better when compared to other studies in the literature and the accuracy rate was 98.90% in the Support Vector Machines (SVM) classifier.Öğe The Role of Vulnerable Software Metrics on Software Maintainability Prediction(Dergipark, 2021) Batur Şahin, CananSoftware maintainability is among the basic quality features of software engineering. Vulnerability prediction is crucial to protect software maintainability from attacks for cybersecurity. Hence, managing vulnerability in an accurate way is an important phase for the efficient prediction of software maintenance. The existing technologies have achieved many good results in vulnerability detection, but no significant results have been obtained on how effective vulnerability metrics for software maintainability prediction is. As far as we know, this paper is the first study that applies the Deep Learning-based Symbiotic Immune Network Model to develop a software maintainability prediction model using vulnerability software metrics. This study proposes a novel methodology capable of discovering software maintainability metrics in open-source software programs efficiently and accurately. The current study also tries to identify vulnerability metrics frequently utilized in software maintainability. In this paper, five commonly employed open-source projects subjected to attacks, such as Mozilla, Linux Kernel, Xen Hypervisor, glibc, and httpd, are used. In the scope of this research, mentioned five open-source software projects were used as datasets, and they were analyzed with their effect on software maintainability prediction. The analysis of the software metrics was performed, and the descriptive statistics of the software metrics were presented. The current research obtained results of software metrics that accurately predicting software maintenance. Furthermore, the experimental findings confirm the effectiveness of the obtained vulnerability metrics for predicting software maintainability. Our experimental results claim that the proposed Deep Learning-based Symbiotic Immune Network Model enables the prediction of software maintainability to be substantially more effective.Öğe Sentiment Analysis of Covid-19 Tweets by using LSTM Learning Model(Ali KARCI, 2021) Karaca, Yunus Emre; Aslan, SerpilSocial media plays an important role in our lives due to the conditions of the age we live. Nowadays, the most popular social media platform that prioritizes meaningful content sharing is Twitter. In Twitter, which produces big data on an unprecedented scale, users have the opportunity to share their own perspectives, feelings, and experiences, as well as examine the opinions of other individuals. The Coronavirus-2019 (Covid-19) disease, transmitted through close contact and small droplets spread by people coughing, sneezing, or speaking, has created social and economic wounds worldwide. As of July 7, 2021, more than 185 million people worldwide have been diagnosed with the New Coronavirus (Covid-19), and approximately 4 million people have died from this infectious disease. This work focuses on the analysis of the sentiments that Covid-19 leaves on people, using the tweets that people share about the Covid-19 pandemic on the Twitter platform. Analyzes are based on deep learning algorithms. Sentiment analysis can provide serious benefits. In this study, we used a Long-short Term Memory (LSTM) based network model. Also, we compared the proposed model other machine learning algorithms: Support Vector Machine (SVM), Naïve Bayes and Logistic Regression. Experimental results show that our proposed method can effectively perform sentiment analysis on the Twitter dataset.Öğe Improving Automated Arabic Essay Questions Grading Based on Microsoft Word Dictionary(Springer Science and Business Media Deutschland GmbH, 2021) Hailat, Muath; Otair, Mohammed; Abualigah, Laith; Houssein, Essam; Batur Şahin, CananThere are three main types of questions: true/false, multiple choice, and essay questions; it is easy to implement automatic grading system (AGS) for multiple choice and true/false questions because the answers are specific compared with essay question answers. Automatic grading system (AGS) was developed to evaluate essay answers using a computer program that solves manual grading process problems like high cost, time-consuming task, increasing number of students, and pressure on teachers. This chapter presents Arabic essay question grading techniques using inner product similarity. The reason behind this is to retrieve students’ answers that more relevance to teachers’ answers. NB (naive Bayes) classifier is used because it is simple to implement and fast. The process starts by preprocessing phase, where tokenization step divides answers for small pieces of tokens. For normalization step, it is used to replace special letter shapes and remove diacritics. Then, stop word removal step removes meaningless and useless words. Finally, stemming process is used to get the stem and root of the words. All the preprocessing phase is meant to be implemented for both student answer and dataset. Then, classifying by naive Bayes classifier to get accurate result also for both students’ answers among with dataset. After that, using Microsoft Word dictionary to compare and get enough synonyms for both students’ answers and model answers in order to have exceptional results. Finally, showing results with the use of inner product similarity then compare the results showed by inner product similarity with human score results so the evaluation among with the efficiency of the proposed technique can be measured using mean absolute error (MAE) and Pearson correlation results (PCR). According to the experimental results, the approach leads to positive results when using MS dictionary and improvement Automated Arabic essay questions grading, where experiment results showed improvement in MAE is 0.041 with enhanced accuracy is 4.65% and PCR is 0.8250. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.Öğe DCW-RNN: Improving Class Level Metrics for Software Vulnerability Detection Using Artificial Immune System with Clock-Work Recurrent Neural Network(IEEE (Institute of Electrical and Electronics Engineers), 2021) Batur Şahin, CananAs the defenses evolve, so do the solutions to a software vulnerability. The primary reason for security incidents, e.g., cyber-attacks, originates from software vulnerabilities. It is challenging to enhance the performance of software processes and determine and eliminate software vulnerabilities. Thus, the development of algorithms with higher security to be applied to possible security issues in software represents a significant research subject for researchers in the domain of software security. The basis of the Dendritic Cell Algorithm (DCA), which is an emerging evolutionary algorithm, constitutes the behavior of specific immune agents, called dendritic cells (DCs). Till now, no strategy or idea has already been adopted on the Clock-Work Recurrent Neural Network (RNN) based Dendritic cell algorithm on vulnerability detection problems. In the present research, the first Clock-Work RNN based Dendritic Cell Algorithm (DCA) was suggested to identify complex dependencies between vulnerable object-oriented software metrics. The suggested method establishes immunity in software vulnerability prediction models to analyze the comparison of the Artificial Immune System Algorithms. The current paper involves the enhanced Clock-Work RNN based Dendritic Cell Algorithm, Genetic Algorithm (GA), and Clonal Selection Algorithm (CLONALG). Furthermore, comparison some studies was made on the basis Artificial Immune System (AIS) algorithms, such as Negative Selection Algorithm (NSA), Cellular Automata (CA), Membrane Computing (P-Systems). The experimental findings of our study demonstrate that our approach was computationally efficient on three different Java projects: Apache Tomcat (releases 6 and 7), Apache CXF, and the Stanford SecuriBench datasets.Öğe Augmented grasshopper optimization algorithm by differential evolution: a power scheduling application in smart homes(Springer, 2021) Ziadeh, Ahmad; Abualigah, Laith; Abd Elaziz, Mohamed; Batur Şahin, Canan; Almazroi, Abdulwahab Ali; Omari, MahmoudWith the increasing number of electricity consumers, production, distribution, and consumption problems of produced energy have appeared. This paper proposed an optimization method to reduce the peak demand using smart grid capabilities. In the proposed method, a hybrid Grasshopper Optimization Algorithm (GOA) with the self-adaptive Differential Evolution (DE) is used, called HGOA. The proposed method takes advantage of the global and local search strategies from Differential Evolution and Grasshopper Optimization Algorithm. Experimental results are applied in two scenarios; the first scenario has universal inputs and several appliances. The second scenario has an expanded number of appliances. The results showed that the proposed method (HGOA) got better power scheduling arrangements and better performance than other comparative algorithms using the classical benchmark functions. Moreover, according to the computational time, it runs in constant execution time as the population is increased. The proposed method got 0.26?% enhancement compared to the other methods. Finally, we found that the proposed HGOA always got better results than the original method in the worst cases and the best cases.Öğe A novel deep learning-based feature selection model for improving the static analysis of vulnerability detection(Springer, 2021) Batur Şahin, Canan; Abualigah, LaithThe automatic detection of software vulnerabilities is considered a complex and common research problem. It is possible to detect several security vulnerabilities using static analysis (SA) tools, but comparatively high false-positive rates are observed in this case. Existing solutions to this problem depend on human experts to identify functionality, and as a result, several vulnerabilities are often overlooked. This paper introduces a novel approach for effectively and reliably finding vulnerabilities in open-source software programs. In this paper, we are motivated to examine the potential of the clonal selection theory. A novel deep learning-based vulnerability detection model is proposed to define features using the clustering theory of the clonal selection algorithm. To our knowledge, this is the first time we have used deep-learned long-lived team-hacker features to process memories of sequential features and mapping from the entire history of previous inputs to target vectors in theory. With an immune-based feature selection model, the proposed approach aimed to improve static analyses' detection abilities. A real-world SA dataset is used based on three open-source PHP applications. Comparisons are conducted based on using a classification model for all features to measure the proposed feature selection methods' classification improvement. The results demonstrated that the proposed method got significant enhancements, which occurred in the classification accuracy also in the true positive rate.
