Datasets:
metadata
license: apache-2.0
task_categories:
- text-classification
language:
- en
tags:
- networking
- wifi
- cybersecurity
- synthetic-data
- connected-home
pretty_name: SmartNet Network Incidents
size_categories:
- 1K<n<10K
SmartNet Network Incidents
Dataset description
SmartNet Network Incidents is a synthetic dataset for experimenting with classification of connected-home networking and security incidents.
Each record combines structured telemetry with a short natural-language incident description.
Supported tasks
- Network incident classification
- Support-ticket triage
- Explainable troubleshooting demonstrations
- Retrieval and RAG experiments
- Model-serving and observability demonstrations
Labels
| Label | Meaning |
|---|---|
healthy |
No material network problem detected |
weak_signal |
Poor Wi-Fi signal or excessive distance/attenuation |
wifi_congestion |
Heavy channel utilization or interference |
high_latency |
Elevated latency or packet loss |
dhcp_failure |
Address allocation or lease-renewal problem |
dns_security_risk |
Suspicious or unusually risky DNS behaviour |
Fields
| Field | Type | Description |
|---|---|---|
incident_id |
string | Synthetic record identifier |
timestamp |
string | Synthetic UTC timestamp |
device_type |
string | Device category |
rssi_dbm |
integer | Received signal strength |
channel_utilization_pct |
integer | Wi-Fi channel utilization |
latency_ms |
integer | Observed network latency |
packet_loss_pct |
float | Estimated packet loss |
dhcp_failures |
integer | Recent DHCP failures |
dns_risk_score |
float | Synthetic DNS risk score |
incident_text |
string | Natural-language incident summary |
label |
string | Target incident category |
Data generation
The records are generated programmatically with deterministic random seeds, label-specific telemetry ranges, varied text templates, and small amounts of noise. No customer records, personal data, or production telemetry are used.
Intended use
This dataset is intended for education, portfolio demonstrations, prototyping, testing pipelines, and comparing classification approaches.
Limitations
- The data is synthetic.
- Labels originate from generation rules rather than human annotation.
- Reported model scores may overestimate real-world performance.
- The dataset does not represent every router, client, environment, or attack.
- It must not be used for safety-critical or production security decisions without validation on representative real-world data.
Privacy
The dataset contains no real users, addresses, identifiers, or customer telemetry.