A clustering system with Siamese Neural Network and Density-based Spatial Clustering (DBSCAN) to group screenshots of malicious web pages based on their semantic similarities so that different sets of similar malicious web pages belonging to different social engineering attack campaigns can be identified.
References
- Lamba, Harshall (2019, Jan 21). “One Shot Learning with Siamese Networks using Keras”
https://towardsdatascience.com/one-shot-learning-with-siamese-networks-using-keras-17f34e75bb3d - Salton do Prado, Kelvin (2017, Apr 1). “How DBSCAN works and why should we use it?”
https://towardsdatascience.com/how-dbscan-works-and-why-should-i-use-it-443b4a191c80 - Rawlani, Himanshu (2020, Jan 1). “Understanding and implementing a fully convolutional network (FCN)”
https://towardsdatascience.com/implementing-a-fully-convolutional-network-fcn-in-tensorflow-2-3c46fb61de3b - Prabhakaran, Selva (2018, Oct 22). “Cosine Similarity - Understanding the math and how it works (with python codes)”
https://www.machinelearningplus.com/nlp/cosine-similarity/ - G Koch, R Zemel, and R Salakhutdinov. “Siamese neural networks for one-shot image recognition”. In ICML Deep Learning workshop, 2015.
- Das, Shibsankar (2019, Jul 17). “Image similarity using Triplet Loss”
https://towardsdatascience.com/image-similarity-using-triplet-loss-3744c0f67973 - Essam, Hazem and Valdarrama, Santiago L (2021, March 25). “Image similarity estimation using a Siamese Network with a triplet loss”
https://keras.io/examples/vision/siamese_network/ - Maklin, Cory (2019, Jun 30). “DBSCAN Python Example: The Optimal Value For Epsilon (EPS)”
https://towardsdatascience.com/machine-learning-clustering-dbscan-determine-the-optimal-value-for-epsilon-eps-python-example-3100091cfbc