--- title: NSG Dataset (Network Security Group Logs) license: mit datasets: - nsg_dataset tags: - cybersecurity - network-security - anomaly-detection - machine-learning --- # NSG Dataset (Network Security Group Logs) ## License **MIT License** ## Dataset Overview The **NSG Dataset** contains **500 records** of simulated **Network Security Group (NSG) logs** to analyze security threats, detect anomalies, and build AI-driven threat prediction models. It includes details on network traffic, threat levels, and response actions. ## Dataset Files - `nsg_dataset.csv` - Contains the main dataset with 500 records. - `README.md` - Documentation for understanding and using the dataset. ## Data Schema | Column Name | Data Type | Description | |-----------------|----------|-------------| | `timestamp` | `string` (YYYY-MM-DD HH:MM:SS) | Date and time of the network event | | `source_ip` | `string` (IPv4) | IP address of the source device | | `destination_ip` | `string` (IPv4) | IP address of the target device | | `protocol` | `string` | Network protocol (TCP, UDP, ICMP) | | `port` | `integer` | Destination port (20 - 65535) | | `action` | `string` | Whether the traffic was **Allowed** or **Denied** | | `threat_level` | `string` | Severity of the security event (**Low, Medium, High, Critical**) | | `threat_type` | `string` | Type of attack (**DDoS, Brute Force, SQL Injection, Port Scan, Malware**) | | `response_action`| `string` | Action taken (**Blocked, Alerted, Monitored, Escalated**) | ## Dataset Statistics - **Total Records:** 500 - **Unique Source IPs:** ~400+ - **Threat Level Distribution:** - Low (~25%) - Medium (~30%) - High (~25%) - Critical (~20%) ## Use Cases - **Anomaly Detection:** Identify unusual traffic patterns. - **Threat Intelligence:** Analyze network security trends. - **Machine Learning Models:** Train AI models for cyber threat prediction. - **Power BI Dashboards:** Visualize security logs for real-time monitoring. ## How to Use the Dataset ### **Python (Pandas) Example** ```python import pandas as pd df = pd.read_csv("nsg_dataset.csv") print(df.head()) # Preview first 5 records