Sentiment Analyzer — Text Signals, Emotion & Keywords

Analyze text sentiment, emotion keywords, polarity, confidence, and evidence snippets for reviews, comments, support tickets, and datasets.

What this actor does

Analyze text sentiment, emotion keywords, polarity, confidence, and evidence snippets for reviews, comments, support tickets, and datasets. The actor is configured for public or user-supplied sources, evidence-backed output rows, capped runs, and KVS summaries for blocked, empty, duplicate, or capped work.

What you can use it for

• Build repeatable sentiment analyzer datasets for analysis and automation.
• Export records with source URLs, evidence fields, and run summaries.
• Run buyer-ready tasks for: Analyze sentiment for customer reviews; Classify support messages by tone; Export sentiment scores for text rows.

Search-intent terms: sentiment analyzer, text sentiment, emotion detection, review sentiment, NLP utility, text analysis.

Inputs

The input schema supports the actor's source fields plus cost and quality controls. Important caps include texts, language, includeEvidence, maxTextCharacters, batchSize. Public task examples below provide small, safe starting inputs.

Outputs

Paid dataset rows and paid KVS artifacts are written only after the configured PPE charge succeeds. Rows include source/evidence fields where the actor can determine them and avoid fabricated values.

Key-value store reports

Every run writes SUMMARY.md and SUMMARY.json. Diagnostic CSV or JSON reports are used for blocked pages, no-result pages, duplicates, unsupported inputs, and capped runs instead of paid dataset rows.

Pricing

This actor uses pay-per-event pricing with Apify platform usage included in the event prices. Store discount tiers are configured for FREE, BRONZE, SILVER, and GOLD users.

Event	Unit	FREE	BRONZE	SILVER	GOLD
actor-start	run	$0.00500	$0.00500	$0.00450	$0.00400
text-analyzed	record	$0.00390	$0.00390	$0.00351	$0.00312

Primary paid output event: text-analyzed. Runtime-paid dataset rows and runtime-paid artifacts use this active paid-output event unless a future pricing revision adds an actor-specific artifact boundary with matching code.

Cost controls and max-charge behavior

Default max charge: $5.00. When the run reaches the max charge or an event-charge limit, it stops gracefully and writes free summary diagnostics without adding unpaid dataset rows.

Platform usage / proxy / browser / API notes

The configured model is fixed-inclusive PPE: platform usage is included in event prices and the platform-usage pass-through toggle is off. Features that require browser rendering, proxies, provider APIs, or additional fetches are bounded by the input caps and priced through explicit events.

Reliability notes and limitations

The actor processes public or permissioned inputs only. Blocked pages, login walls, unsupported formats, empty results, and provider errors are recorded as diagnostics. The actor does not fabricate rankings, prices, reviews, sentiment, traffic, compliance status, or private contact data when source evidence is unavailable.

Example inputs

{
  "texts": [
    "This product is amazing! Best purchase I've ever made.",
    "Terrible service, waited 2 hours and nobody helped.",
    "The package arrived on Tuesday as expected."
  ],
  "detectEmotions": true,
  "extractKeyPhrases": true,
  "detectLanguage": true,
  "sentimentModel": "combined",
  "neutralThreshold": 0.05,
  "includeWordScores": false,
  "onlySentiment": "positive",
  "minimumConfidence": 0,
  "maxChargeUsd": 5,
  "debugMode": false,
  "saveRawHtmlToKvs": false,
  "requestTimeout": 30000,
  "maxRetries": 1,
  "maxConcurrency": 1,
  "language": "en",
  "includeEvidence": "sentiment analyzer",
  "maxTextCharacters": 1,
  "batchSize": "sentiment analyzer",
  "sourceMode": "livePublic",
  "suppliedSnapshotHtml": "sentiment analyzer",
  "suppliedSnapshotJson": "sentiment analyzer",
  "suppliedSnapshotText": "sentiment analyzer"
}

Public task examples

• Analyze sentiment for customer reviews: Dataset rows matching the analyze sentiment for customer reviews use case, plus SUMMARY.md and SUMMARY.json reports in the key-value store.
• Classify support messages by tone: Dataset rows matching the classify support messages by tone use case, plus SUMMARY.md and SUMMARY.json reports in the key-value store.
• Export sentiment scores for text rows: Dataset rows matching the export sentiment scores for text rows use case, plus SUMMARY.md and SUMMARY.json reports in the key-value store.

Ethical/allowed-use notes

Use this actor only for public data, owned files, or sources you have permission to process. Respect source terms, privacy expectations, robots guidance, and applicable laws. Do not use the actor to collect secrets, bypass access controls, or make unsupported legal, financial, medical, or compliance decisions.