Analyze the sensitivity level of one or multiple keywords, returning standardized risk labels and confidence results.

Overview

- Risk Analysis: Provides semantic-level risk judgments based on text content.

- Stable Response: Balances processing efficiency and response speed for high-frequency call scenarios.

- Concurrent Support: Supports batch concurrent processing, can analyze up to 100 keywords at once.

- Input Limits: Single keyword max 1,000 characters, total characters max 20,000.

- Standard Labels: Returns label field, clearly distinguishing between sensitive and normal.

- Clear Classification: Returns category field to identify specific risk categories.

- Confidence Output: Returns confidence field, ranging from 0.0 to 1.0.

Response Field Descriptions

| Field | Type | Description |

|------|------|-------------|

| results | array | Array of analysis result objects. |

| results[].k | string | The original keyword you provided in the request. |

| results[].label | string | Core judgment field: sensitive, normal. |

| results[].category | string | Risk classification: safe, threat, porn, fraud, insult. |

| results[].confidence | number | Confidence of current classification, range 0.0 to 1.0. |

| total | integer | Total number of keywords successfully analyzed in this request.