Stimmungsanalyse

Beschreibung

Analyze a given text to determine its sentiment, providing a detailed breakdown of positive, negative, and neutral elements, along with an overall sentiment score and analysis of undertones.

Projektkompatibilität

Windows | Plattformübergreifend

Konfiguration

• Verbindungs-ID – Die im Integration Service hergestellte Verbindung. Greifen Sie auf das Dropdownmenü zu, um Verbindungen auszuwählen, hinzuzufügen oder zu verwalten.

• Text - The text to be analyzed for sentiment. This field supports String type input.

Eigenschaften verwalten

Verwenden Sie den Assistenten Eigenschaften verwalten , um Standard- oder benutzerdefinierte Felder des Objekts zu konfigurieren oder zu verwenden. Sie können Felder auswählen, um sie der Aktivitäts-Canvas hinzuzufügen. Die hinzugefügten Standard- oder benutzerdefinierten Felder sind im Eigenschaftenbereich (in Studio Desktop) oder unter Zusätzliche Eigenschaften anzeigen (in Studio Web) verfügbar.

Zusätzliche Eigenschaften

Ausgabe

• Overall Sentiment - Contains the sentiment score and label:
  • Very Negative: -99 to -67
  • Negative: -66 to -34
  • Slightly Negative: -33 to -1
  • Neutral: 0
  • Slightly Positive: 1 to 33
  • Positive: 34 to 66
  • Very Positive: 67 to 99
• Sentiment Breakdown - Counts of positive, negative, neutral, and total statements.
• Confidence Level - The overall confidence level of the analysis.
• Key Phrases - A string of key phrases identified in the sentiment analysis.
• Analysis - Detailed explanation of the sentiment analysis.
• Undertones - Subtle undertones detected in the text with their impact.
• Sentiment Analysis - Automatically generated output variable.

Output JSON format:

{
  "overallSentiment": {
    "score": 0,
    "label": ""
  },
  "sentimentBreakdown": {
    "positiveStatements": 0,
    "negativeStatements": 0,
    "neutralStatements": 0,
    "totalStatements": 0
  },
  "confidenceLevel": 0,
  "keyPhrases": [
    {
      "phrase": "",
      "sentiment": "",
      "confidence": 0
    }
  ],
  "analysis": "",
  "undertones": [
    {
      "description": "",
      "impact": ""
    }
  ]
}

Limitations and other considerations

• The complexity and length of the input text may affect the accuracy of the analysis.

• The system may have difficulties with highly context-dependent or culturally specific expressions.

• Sarcasm, irony, and other forms of figurative language may pose challenges for accurate sentiment detection.

• The system's performance may vary across different languages or dialects.