Relevance Checking

Verify that comments stay on-topic and don't violate banned topic rules.

What it checks:

• On-topic detection: Is the comment relevant to the article?
• Banned topics: Does it discuss topics you've prohibited?

Reference

See the Relevance Checking API Reference for method signatures and parameters.

Response: CommentRelevanceResult

By Example

Basic Usage

TypeScript

const result = await client.checkRelevance(
  "This comment discusses the article topic",
  articleId
);

console.log(`On topic: ${result.on_topic.on_topic}`);
console.log(`Confidence: ${result.on_topic.confidence.toFixed(2)}`);
console.log(`Reasoning: ${result.on_topic.reasoning}`);

// Banned topics info
console.log(`Banned topics found: ${result.banned_topics.banned_topics}`);
console.log(`Quantity on banned: ${result.banned_topics.quantity_on_banned_topics.toFixed(2)}`);

Python

Blocking Client

result = client.check_relevance(
    "This comment discusses the article topic",
    article_id
)

print(f"On topic: {result.on_topic.on_topic}")
print(f"Confidence: {result.on_topic.confidence:.2f}")
print(f"Reasoning: {result.on_topic.reasoning}")

# Banned topics info
print(f"Banned topics found: {result.banned_topics.banned_topics}")
print(f"Quantity on banned: {result.banned_topics.quantity_on_banned_topics:.2f}")

Async Client

result = await client.check_relevance(
    "This comment discusses the article topic",
    article_id
)

print(f"On topic: {result.on_topic.on_topic}")
print(f"Confidence: {result.on_topic.confidence:.2f}")
print(f"Reasoning: {result.on_topic.reasoning}")

# Banned topics info
print(f"Banned topics found: {result.banned_topics.banned_topics}")
print(f"Quantity on banned: {result.banned_topics.quantity_on_banned_topics:.2f}")

PHP

$client->checkRelevance($articleId, "This comment discusses the article topic")
    ->then(function ($result) {
        echo "On topic: " . ($result->onTopic->onTopic ? 'Yes' : 'No') . "\n";
        echo "Confidence: " . number_format($result->onTopic->confidence, 2) . "\n";
        echo "Reasoning: " . $result->onTopic->reasoning . "\n";

        echo "Banned topics: " . implode(', ', $result->bannedTopics->bannedTopics) . "\n";
    });

$client->run();

REST API

Try it out live →

curl -X POST https://app.respectify.ai/v0.2/commentrelevance \
  -H "X-User-Email: your@email.com" \
  -H "X-API-Key: your-api-key" \
  -d "article_context_id=550e8400-e29b-41d4-a716-446655440000" \
  -d "comment=This comment discusses the article topic"

Response:

{
  "on_topic": {
    "reasoning": "The comment is directly related to the article content.",
    "on_topic": true,
    "confidence": 0.9
  },
  "banned_topics": {
    "reasoning": "No banned topics detected.",
    "banned_topics": [],
    "quantity_on_banned_topics": 0,
    "confidence": 1.0
  }
}

With Banned Topics

Specify topics that shouldn't be discussed:

TypeScript

const result = await client.checkRelevance(
  "Let me tell you about politics...",
  articleId,
  ["politics", "religion", "controversial-subject"]
);

if (result.banned_topics.banned_topics.length > 0) {
  console.log(`Comment violates banned topics: ${result.banned_topics.banned_topics}`);
}

Python

Blocking Client

result = client.check_relevance(
    "Let me tell you about politics...",
    article_id,
    banned_topics=['politics', 'religion', 'controversial-subject']
)

if result.banned_topics.banned_topics:
    print(f"Comment violates banned topics: {result.banned_topics.banned_topics}")

Async Client

result = await client.check_relevance(
    "Let me tell you about politics...",
    article_id,
    banned_topics=['politics', 'religion', 'controversial-subject']
)

if result.banned_topics.banned_topics:
    print(f"Comment violates banned topics: {result.banned_topics.banned_topics}")

PHP

$client->checkRelevance(
    $articleId,
    "Let me tell you about politics...",
    ['politics', 'religion', 'controversial-subject']
)->then(function ($result) {
    if (!empty($result->bannedTopics->bannedTopics)) {
        echo "Violates: " . implode(', ', $result->bannedTopics->bannedTopics);
    }
});

REST API

Try it out live →

curl -X POST https://app.respectify.ai/v0.2/commentrelevance \
  -H "X-User-Email: your@email.com" \
  -H "X-API-Key: your-api-key" \
  -d "article_context_id=550e8400-e29b-41d4-a716-446655440000" \
  -d "comment=Let me tell you about politics..." \
  -d "banned_topics[]=politics" \
  -d "banned_topics[]=religion" \
  -d "banned_topics[]=controversial-subject"

Response:

{
  "on_topic": {
    "reasoning": "The comment veers off-topic from the article's subject matter.",
    "on_topic": false,
    "confidence": 0.85
  },
  "banned_topics": {
    "reasoning": "Comment discusses politics, which is in the banned topics list.",
    "banned_topics": ["politics"],
    "quantity_on_banned_topics": 0.8,
    "confidence": 0.92
  }
}