Index of supplementary files from "Perils of Zero-Interaction Security in the Internet of Things"

This record serves an an index to the other dataset releases that are part of the paper "Perils of Zero Interaction Security in the Internet of Things" by Mikhail Fomichev, Max Maass, Lars Almon, Alejandro Molina, Matthias Hollick, in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 3, Issue 1. We have chosen to split the dataset into several parts to meet Zenodo size requirements and make it easier to find specific pieces of data. In total, the following datasets exist: Raw data These datasets contain raw data, as collected directly from the devices doing the recording. It includes readings from several different sensors, as well as observed WiFi and BLE signals with their signal strength, and in one case, audio recordings. This raw data can be used to repeat our own experiments, or to apply different schemes to it to have a baseline for comparisons. Four datasets exist, mapped to the three scenarios discussed in the paper: Car Scenario Office Scenario Mobile Scenario + audio data in separate deposit (with access control) Processed Data The processed data is generated from the raw data using the processing code (which can be found in the code repository). The resulting data contains computed features from the five papers under investigation plus derived machine learning datasets, and can be used to see in detail how the schemes behave in specific situations. These datasets tend to be fairly large. Three datasets exist: Car Scenario Office Scenario Mobile Scenario Result Data Finally, the result datasets contain the results of the evaluation (i.e., the computed error rates and generated plots, plus associated caches). The code used to derive these results can once again be found in the source code repository. Here, five datasets exist, one for each investigated paper: Karapanos et al. Schürmann and Sigg Miettinen et al. Truong et al. Shrestha et al.

本记录作为论文《物联网零交互安全的隐患》（Perils of Zero Interaction Security in the Internet of Things）配套数据集发布项的索引，该论文作者为Mikhail Fomichev、Max Maass、Lars Almon、Alejandro Molina、Matthias Hollick，发表于《ACM交互式、移动、可穿戴与普适技术汇刊》（Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies）第3卷第1期。为适配Zenodo的文件大小限制并便于用户精准检索特定数据片段，我们将整体数据集拆分为多个子部分。目前共包含以下三类数据集：

### 原始数据（Raw data）
此类数据集收录直接从采集设备获取的原始采集结果，涵盖多种传感器的读数、已观测到的WiFi与蓝牙低功耗（BLE, Bluetooth Low Energy）信号及其信号强度，其中一个数据片段还包含音频录制内容。该原始数据可用于复现本研究的实验，或应用不同算法以获取用于对比的基准结果。本次共发布四份原始数据集，对应论文中提及的三个场景：车载场景、办公场景、移动场景，另有一份包含音频数据的存档需单独申请访问权限。

### 处理后数据（Processed Data）
处理后数据通过源代码仓库中的处理脚本从原始数据生成，包含所调研的五篇论文对应的计算特征与衍生机器学习数据集，可用于细致分析各类算法在特定场景下的运行表现。此类数据集通常体积较大，本次共发布三份处理后数据集，分别对应车载场景、办公场景与移动场景。

### 结果数据（Result Data）
结果数据集包含评估阶段生成的最终结果，即计算得到的错误率、生成的可视化绘图文件及相关缓存。用于生成此类结果的代码同样可在源代码仓库中获取。本次共发布五份结果数据集，分别对应每一篇被调研的论文：Karapanos等人的研究、Schürmann与Sigg的研究、Miettinen等人的研究、Truong等人的研究以及Shrestha等人的研究。