Topical Behavior Prediction from Massive Logs
This work addresses behavior prediction for entities in network logs, but it appears incremental as it builds on existing deep learning methods with specific adaptations.
The paper tackles the problem of predicting topical behavior from large-scale network logs by exploring temporal and spatial relationships using deep learning architectures combining RNN and CNN, and introduces a new framework called SCCN, showing gains compared to an MLP network.
In this paper, we study the topical behavior in a large scale. We use the network logs where each entry contains the entity ID, the timestamp, and the meta data about the activity. Both the temporal and the spatial relationships of the behavior are explored with the deep learning architectures combing the recurrent neural network (RNN) and the convolutional neural network (CNN). To make the behavioral data appropriate for the spatial learning in the CNN, we propose several reduction steps to form the topical metrics and to place them homogeneously like pixels in the images. The experimental result shows both temporal and spatial gains when compared against a multilayer perceptron (MLP) network. A new learning framework called the spatially connected convolutional networks (SCCN) is introduced to predict the topical metrics more efficiently.