README.md

---
license: cc-by-nc-4.0
task_categories:
- tabular-classification
- time-series-forecasting
tags:
- tsfile
- timeseries
- cybersecurity
- soc-operations
- alert-triage
- mitre-attack
- soar
- siem
- synthetic-data
- incident-response
- analyst-fatigue
- false-positive-reduction
- time-series
- format:tsfile
pretty_name: CYB008 — Synthetic SOC Alert Dataset (Sample) (TsFile)
size_categories:
- 10K<n<100K
configs:
- config_name: default
  data_files:
  - split: alert_events
    path: alert_events.tsfile
  - split: incident_summary
    path: incident_summary.tsfile
  - split: soc_alerts
    path: soc_alerts.tsfile
  - split: soc_topology
    path: soc_topology.tsfile
---

# CYB008 — Synthetic SOC Alert Dataset (Sample) (TsFile)

Apache TsFile version of [`xpertsystems/cyb008-sample`](https://huggingface.co/datasets/xpertsystems/cyb008-sample).

## Overview

**CYB008 — Synthetic SOC Alert Dataset (Sample)** — This is a **free preview** of the full **CYB008 — Synthetic SOC Alert Dataset** product. It contains roughly **~10% of the full dataset** at identical schema, MITRE ATT&CK tactic coverage, and statistical fingerprint, so you can evaluate fit before licensing the full product.

- **Tables:** `alert_events`, `incident_summary`, `soc_alerts`, `soc_topology`
- **Rows:** 65,112 across 4 TsFile table(s).
  - `alert_events`: 55,298 rows — Time is the source `event_timestamp_min` simulation clock in **minutes**, converted to milliseconds (×60,000).
  - `incident_summary`: 589 rows — Time is the source `first_alert_timestamp_min` simulation clock in **minutes**, converted to milliseconds (×60,000).
  - `soc_alerts`: 9,200 rows — Time is the source `alert_timestamp_min` simulation clock in **minutes**, converted to milliseconds (×60,000).
  - `soc_topology`: 25 rows — Time is synthetic row index over the stable source order (0 … N-1); the source table has no per-row clock.

## Schema (TsFile structure)

### `alert_events.tsfile`

- **Time** (INT64, milliseconds) — the source `event_timestamp_min` simulation clock in **minutes**, converted to milliseconds (×60,000).
- **TAG**: `analyst_id` — device dimension(s); query one device with `WHERE analyst_id=...`.
- **FIELD** (9): `event_id` (string), `alert_id` (string), `event_type` (string), `alert_severity` (string), `mitre_tactic` (string), `detection_source_id` (string), `escalation_target_tier` (string), `soar_playbook_id` (string), `partner_alert_id` (string)

### `incident_summary.tsfile`

- **Time** (INT64, milliseconds) — the source `first_alert_timestamp_min` simulation clock in **minutes**, converted to milliseconds (×60,000).
- **TAG**: `incident_id` — device dimension(s); query one device with `WHERE incident_id=...`.
- **FIELD** (21): `soc_id` (int64), `analyst_id_lead` (int64), `incident_severity` (string), `incident_type` (string), `alert_count_linked` (int64), `incident_declared_timestamp_min` (double), `incident_closed_timestamp_min` (double), `mttd_minutes` (double), `mttr_minutes` (double), `false_positive_rate` (double), `escalation_count` (int64), `soar_actions_taken` (int64), `analyst_tier_peak` (string), `kill_chain_stages_observed` (int64), `mitre_tactics_covered` (string), `containment_action_count` (int64), `threat_hunt_triggered` (int64), `incident_success_flag` (int64), `analyst_fatigue_score` (double), `sla_breached_flag` (int64), `auto_resolution_fraction` (double)

### `soc_alerts.tsfile`

- **Time** (INT64, milliseconds) — the source `alert_timestamp_min` simulation clock in **minutes**, converted to milliseconds (×60,000).
- **TAG**: `soc_id` — device dimension(s); query one device with `WHERE soc_id=...`.
- **FIELD** (25): `alert_id` (string), `incident_id` (string), `analyst_id` (int64), `shift_id` (string), `alert_lifecycle_phase` (string), `alert_severity` (string), `alert_source` (string), `mitre_tactic` (string), `mitre_technique_id` (string), `detection_rule_id` (string), `raw_score` (double), `enriched_score` (double), `analyst_tier` (string), `time_in_phase_minutes` (double), `queue_depth_at_ingestion` (int64), `soar_playbook_triggered` (int64), `automation_resolved` (int64), `escalation_flag` (int64), `sla_breached_flag` (int64), `resolution_outcome` (string), `mttd_minutes` (double), `mttr_minutes` (double), `fatigue_score_at_alert` (double), `siem_platform` (string), `soar_playbook_id` (string)

### `soc_topology.tsfile`

- **Time** (INT64, milliseconds) — synthetic row index over the stable source order (0 … N-1); the source table has no per-row clock.
- **TAG**: `soc_id` — device dimension(s); query one device with `WHERE soc_id=...`.
- **FIELD** (12): `analyst_id` (int64), `analyst_tier` (string), `shift_pattern` (string), `siem_platform` (string), `detection_source_type` (string), `automation_coverage_fraction` (double), `soar_playbook_count` (int64), `detection_rule_count` (int64), `alert_volume_daily_mean` (int64), `false_positive_rate_baseline` (double), `escalation_policy` (string), `max_concurrent_alerts` (int64)

## Conversion notes

- Each CSV of the sample is converted into its own TsFile table (4 tables), so no table is merged with another and no column is dropped.
- Event/telemetry tables use the source simulation clock as Time (INT64 milliseconds): `timestep` step counters and millisecond offsets are stored as-is, minute-based columns are multiplied by 60,000, and ISO/UTC strings are parsed to epoch milliseconds.
- Static inventory/topology/summary tables have no per-row clock; their Time is a synthetic row index (0..N-1) over the stable source order.
- Low-cardinality dimensions of each table (campaign/incident/asset/session ids, phases, labels) are declared as TAG columns; query one simulation with e.g. `WHERE campaign_id='...'`.

## Usage

Install the Apache TsFile Python SDK (`pip install tsfile`) and read a converted file:

```python
from pathlib import Path
from tsfile import TsFileReader

path = Path("alert_events.tsfile")
with TsFileReader(str(path)) as reader:
    schemas = reader.get_all_table_schemas()
    print("tables:", list(schemas))
    table_name = next(iter(schemas))
    table = schemas[table_name]
    columns = [column.get_column_name() for column in table.get_columns()]
    print("columns:", columns)
    field_names = [
        column.get_column_name()
        for column in table.get_columns()
        if column.get_column_name() not in {"Time", "time"}
    ]
    if field_names:
        with reader.query_table(table_name, field_names[:3], batch_size=1024) as result:
            batch = result.read_arrow_batch()
            if batch is not None:
                print(batch.to_pandas().head())
```

## Source & license

- Original dataset: [`xpertsystems/cyb008-sample`](https://huggingface.co/datasets/xpertsystems/cyb008-sample)
- Source card title / publisher: CYB008 — Synthetic SOC Alert Dataset (Sample)
- License: cc-by-nc-4.0 (as declared by the original dataset).