Summary Uniform resource locator (URL)‐based cyber‐attacks form a major part of security threats in cyberspace. Even though the experience and awareness of the end‐users help them protect themselves from these attacks, a software‐based solution is necessary for comprehensive protection. To this end, a novel robust URL classification model based on convolutional neural network is proposed in this study. The proposed model classifies given URLs into five classes, namely, () , () , () , () , and () . The proposed model was trained and evaluated on a gold standard URL dataset comprising of samples. According to the experimental result, the proposed model obtained an accuracy as high as which outperformed the state‐of‐the‐art. Based on the same architecture, we proposed another classifier, a binary classifier that detects malicious URLs without dealing with their types. This binary classifier obtained an accuracy as high as which outperformed the state‐of‐the‐art as well. The experimental result demonstrates the feasibility of the proposed solution.

Abstract The whole world has been experiencing the COVID‐19 pandemic since December 2019. During the pandemic, a new life has been started by necessity where people have extensively used social media to express their feelings, and find information. Twitter was used as the source of what people have shared regarding the COVID‐19 pandemic. Sentiment analysis deals with the extraction of the sentiment of a given text. Most of the related works deal with sentiment analysis in English, while studies for Turkish sentiment analysis lack in the research field. To this end, a novel sentiment analysis model based on the combination of convolutional neural network and bidirectional long short‐term memory was proposed in this study. The proposed deep neural network model was trained on the constructed Twitter dataset, which consists of Turkish tweets regarding the COVID‐19 pandemic, to classify a given tweet into three sentiment classes, namely, (i) , (ii) , and (iii) . A set of experiments were conducted for the evaluation of the proposed model. According to the experimental result, the proposed model obtained an accuracy as high as , which outperformed the state‐of‐the‐art baseline models for sentiment analysis of tweets in Turkish.

Summary The easiest way to propose a solution based on deep neural networks is using the pre‐trained models through the transfer‐learning technique. Deep learning platforms provide various pre‐trained deep neural networks that can be easily applied for image classification tasks. So, “Which pre‐trained model provides the best performance for image classification tasks?” is a question that instinctively comes to mind and should be shed light on by the research community. To this end, we propose an experimental comparison of the six popular pre‐trained deep neural networks, namely, (i) VGG19, (ii) ResNet50, (iii) DenseNet201, (iv) MobileNetV2, (v) InceptionV3, and (vi) Xception by employing them through the transfer‐learning technique. Then, the proposed benchmark models were both trained and evaluated under the same configurations on two gold‐standard datasets, namely, (i) CIFAR‐10 and (ii) Stanford Dogs to benchmark them. Three evaluation metrics were employed to measure performance differences between the employed pre‐trained models as follows: (i) Accuracy, (ii) training duration, and (iii) inference time. The key findings that were obtained through the conducted a wide variety of experiments were discussed.

Abstract Convolutional neural networks have provided state‐of‐the‐art solutions for many subfields of computer vision. While there exist many studies in the literature for several languages, studies for handwritten Turkish character recognition lack in the research field. To this end, we propose a novel handwritten Turkish letter recognition model based on a convolutional neural network. Since, to the best of our knowledge, there do not exist any publicly available handwritten Turkish letters datasets, we constructed a handwritten Turkish letters dataset that consists of 25,875 samples. To compare the performance of the proposed model with the related work, three state‐of‐the‐art models, namely, VGG 19, InceptionV 3, and Xception , were utilized through the transfer learning technique. When these models were evaluated on the handwritten Turkish letter dataset, the proposed model's accuracy was calculated as high as 96.07% which was higher than the benchmark models. To measure the generalization ability of the proposed model, it was evaluated on a gold standard dataset, namely, EMNIST , and has achieved an accuracy of 80.54% which was higher than the benchmark models. Finally, the proposed model was trained and evaluated on the EMNIST dataset and it has achieved an accuracy of 94.61% which outperformed the related work.

As a natural consequence of offering many advantages to their users, social media platforms have become a part of daily lives. Recent studies emphasize the necessity of an automated way of detecting the offensive posts in social media since these ‘toxic’ posts have become pervasive. To this end, a novel toxic post detection approach based on Deep Neural Networks was proposed within this study. Given that several word embedding methods exist, we shed light on which word embedding method produces better results when employed with the five most common types of deep neural networks, namely, , , , , and a combination of and . To this end, the word vectors for the given comments were obtained through four different methods, namely, () , () , () , and () the layer of deep neural networks. Eventually, a total of twenty benchmark models were proposed and both trained and evaluated on a gold standard dataset which consists of tweets. According to the experimental result, the best , , was obtained on the proposed model without employing pre-trained word vectors which outperformed the state-of-the-art works and implies the effective embedding ability of s. Other key findings obtained through the conducted experiments are that the models, that constructed word embeddings through the layers, obtained higher s and converged much faster than the models that utilized pre-trained word vectors.

Web 2.0 technologies have not only raised microblogs, but also social software development and collaboration platforms. GitHub is the most popular software development platform that provides social collaboration. Within the scope of this study, a novel graph-based analysis model is proposed which targets to reveal (1) the characteristics of the GitHub in order to shed light on social software development in general, and (2) the most popular programming languages, repositories, and developers in order to shed light on the trending software development technologies. To this end, a subset of the GitHub network, which contains 84, 737 developers and 209, 100 repositories, was collected through the GitHub API and stored on a graph database namely neo4j to be later analyzed. The result of the analysis shows that (1) the connections in GitHub are not mutually linked, (2) JavaScript, Python, and Java are currently the most popular three programming languages, (3) You-Dont-Know-JS, oh-my-zsh, and public-apis are the most popular three repositories, and (4) TarrySingh (Tarry Singh), indrajithban-dara (Indrajith Bandara), and rootsongjc (Jimmy Song) are the most popular three developers. Furthermore, the proposed novel analysis model can be easily applied to other social networks.