The CIS Department is a power in student research at Brooklyn College. CIS had two winners at Science Day in the Undergraduate Division, both of whom tied for second place: Aleksandr Kovalev, mentored by Assistant Professor Tzipora Halevi, on Android malware detection, and Joshua N. Sash, mentored by Assistant Professor Hui Chen, on determining the accuracy of code-similarity programs. Aleksandr Kovalev Alex Kovalev Alex Kovalev presented his work on use of neural networks to detect malicious Android applications. The project introduced a newly designed neural network for detecting malware, based on the features extracted from Android .apk files. The proposed Neural Network (NN) is a supervised feed-forward network with two hidden layers. The algorithm was tested against two existing datasets. Multiple tests were run where the data was split into 66 percent training data and 33 percent testing data. The NN achieved a test accuracy of about 95 percent consistently on the dataset and the recall was in the range of 92 to 94 percent. The classification results were compared to the ones obtained from a previous algorithm (Drebin) and the strengths and weaknesses of both methods were analyzed. Alex was mentored by Assistant Professor Tzipora Halevi. Joshua N. Sash Joshua N. Sash Joshua N. Sash showcased his work, “Determining Accuracy of Code Similarity Programs,” in a poster presentation. He examined a few code similarity programs with potential applications in automated plagiarism detection, for which he devised two strategies, comparing program code directly, and comparing changsets obtained from a source code management system, applied them to program code and revision histories of about 100 students, and obtained encouraging results. Joshua’s work was advised by Assistant Professor Hui Chen.
