社交软件网络流量数据集

本数据集描述了基于深度高斯过程的流量预测数据，主要测量两位用户在使用三种社交软件(ICQ，Skype，Facebook)时所产生的具体网络流量。本数据集的来源为VPN-nonVPN(ISCXVPN2016)，经过数据聚集之后得到。这一数据集在构建时，首先创建了两个用户Alice和Bob，然后在真实的网络场景下，测量两位用户在使用三种社交软件(ICQ，Skype，Facebook)时所产生的具体网络流量。该数据集一共包含11次聊天通讯记录所产生的流量，每次通话根据时长差异，记录包含10至2000个记录点，合计大约5500个数值记录点。

This dataset focuses on network traffic prediction data based on Deep Gaussian Processes, which measures the specific network traffic generated by two users when using three social applications: ICQ, Skype, and Facebook. This dataset is derived from the VPN-nonVPN (ISCXVPN2016) dataset after data aggregation. During its construction, two test users Alice and Bob were created, and their real network traffic generated while utilizing the three aforementioned social platforms was measured in actual network scenarios. The dataset contains traffic generated from a total of 11 chat communication sessions. Depending on the duration of each session, each contains between 10 and 2000 recording points, with an overall total of approximately 5500 numerical recording points.

搜集汇总

背景与挑战

背景概述

该数据集是基于深度高斯过程预测的社交软件网络流量数据，来源于ISCXVPN2016，测量了两位用户使用ICQ、Skype和Facebook三种社交软件产生的具体网络流量。数据集包含11次聊天通讯记录，约5500个数值记录点，数据量为13.39MB，由清华大学深圳研究生院作为国家重点研发计划项目的一部分创建，适用于网络测量和社交媒体相关研究。