> For the complete documentation index, see [llms.txt](https://docs.elastiflow.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.elastiflow.com/additional-resources-reference-articles/faq/netintel-threat-scoring-and-ranking.md).

# NetIntel Threat Scoring and Ranking

## NetIntel Threat Scoring & Ranking

### Overview

NetIntel identifies a fixed catalog of **16 threat types** on each flow and produces three parallel score arrays — **confidence**, **risk**, and **threat** — that let you rank how dangerous each detected flow is. The **threat score** is derived directly from the risk and confidence scores, so it can be computed (and re-derived) at any time:

```
threat.score = risk.score × confidence.score
```

Flows whose threat score lands in the **Critical** band are the highest priority and should be surfaced first.

### Data Model

For each flow NetIntel detects, it emits four **parallel arrays** that are aligned position-by-position. All share the `flow.*.sec` parent prefix:

| Field                                | Contents                                |
| ------------------------------------ | --------------------------------------- |
| `flow.*.sec.threat.name`             | Detected threat names                   |
| `flow.*.sec.threat.confidence.score` | Confidence score per threat (1–3)       |
| `flow.*.sec.threat.risk.score`       | Risk score per threat (1–7)             |
| `flow.*.sec.threat.score`            | Threat score per threat (derived, 1–21) |

The *n*th element of each array describes the **same** threat:

#### Threat Catalog (all 16 types)

The full set of threat types NetIntel can identify, in canonical order:

|  #  | Threat Name               |
| :-: | ------------------------- |
|  1  | SIP Access                |
|  2  | VNC Exploit               |
|  3  | Remote Desktop Access     |
|  4  | HTTP Exploit              |
|  5  | SMTPS Access              |
|  6  | HTTPS Access              |
|  7  | MySQL Database Exploit    |
|  8  | Kamstrup Access           |
|  9  | IMAPS Exploit             |
|  10 | Internet Scanner          |
|  11 | IMAP Exploit              |
|  12 | SOCK5 Access              |
|  13 | Remote Desktop Exploit    |
|  14 | SOCK5 Exploit             |
|  15 | IMAPS Access              |
|  16 | Postgres Database Exploit |

### Score Scales

#### Confidence (how sure NetIntel is about the detection) — range **1–3**

| Value | Meaning           |
| :---: | ----------------- |
|   1   | Low confidence    |
|   2   | Medium confidence |
|   3   | High confidence   |

#### Risk (how dangerous the threat is if real) — range **1–7**

| Value | Meaning   |
| :---: | --------- |
|   1   | Low risk  |
|   …   | …         |
|   7   | High risk |

### Deriving the Threat Score

For each detected threat at position *n*:

```
flow.*.sec.threat.score = flow.*.sec.threat.risk.score × flow.*.sec.threat.confidence.score
```

* **Minimum:** `1 × 1 = 1`
* **Maximum:** `7 × 3 = 21`

So every threat score falls in the range **1–21**.

#### Worked Example

On a given flow, NetIntel detects **VNC Exploit** and reports `risk.score = 7` with `confidence.score = 2` for that entry:

```
threat.score = 7 × 2 = 14  →  Critical
```

That flow would be flagged as the highest priority.

### Threat Severity Bands

| Threat Score Range | Severity     |
| :----------------: | ------------ |
|         1–2        | **Low**      |
|         3–6        | **Medium**   |
|        7–12        | **High**     |
|        14–21       | **Critical** |

### Threat Ranking Grid

*Text version of the "Threat Ranking" chart.* Confidence increases **down the rows**; risk ("danger") increases **across the columns**. Each cell shows `score · severity`.

| Confidence ↓ \ Risk → |   Low (1)  |   Med (3)  |    High (7)   |
| --------------------- | :--------: | :--------: | :-----------: |
| **Low (1)**           |   1 · Low  | 3 · Medium |    7 · High   |
| **Med (2)**           |   2 · Low  | 6 · Medium | 14 · Critical |
| **High (3)**          | 3 · Medium |  9 · High  | 21 · Critical |

**Legend:** `1–2` Low · `3–6` Medium · `7–12` High · `14–21` Critical

#### Full Threat Score Matrix (all risk values)

The grid below expands the chart to every risk value (1–7). Use it to look up any combination quickly.

| Confidence ↓ \ Risk → |    1    |    2    |     3    |     4     |     5     |     6     |     7     |
| --------------------- | :-----: | :-----: | :------: | :-------: | :-------: | :-------: | :-------: |
| **1 (Low)**           | 1 · Low | 2 · Low |  3 · Med |  4 · Med  |  5 · Med  |  6 · Med  |  7 · High |
| **2 (Med)**           | 2 · Low | 4 · Med |  6 · Med |  8 · High | 10 · High | 12 · High | 14 · Crit |
| **3 (High)**          | 3 · Med | 6 · Med | 9 · High | 12 · High | 15 · Crit | 18 · Crit | 21 · Crit |

### How to Use It for Prioritization

* **Filter first on `flow.*.sec.threat.score`:** pull flows where the threat score is in the **Critical (14–21)** range — these are the most important and should be highlighted.
* **Identify the threat type** via the matching position in `flow.*.sec.threat.name` to know *what* is being detected.

### Tooling

The score fields are available in Elasticsearch as well as OpenSearch, so you can build alerts and highlighting on `flow.*.sec.threat.score` directly. A worked walkthrough of automating network alerts (e.g., Elastiflow → Slack) is available at:

* <https://www.elastiflow.com/blog/posts/automate-your-network-alerts-a-step-by-step-elastiflow-and-slack-tutorial>
