Share via

Text Translator - Sentiment Analysis

Lalith 161 Reputation points
2021-10-27T06:55:16.257+00:00

Hi, please look at the json response object. my document has the following text - "The food and service were unacceptable, but the concierge were nice" why is this text considered to be positive? I think it should be negative? please help me understand this.

"sentences": [
        {
          "sentiment": "positive",
          "confidenceScores": {
            "positive": 0.84,
            "neutral": 0.0,
            "negative": 0.16
          },
          "offset": 0,
          "length": 67,
          "text": "The food and service were unacceptable, but the concierge were nice"
        }
      ],
Azure Language in Foundry Tools
Azure Language in Foundry Tools

An Azure service that provides natural language capabilities including sentiment analysis, entity extraction, and automated question answering.

Foundry Tools
Foundry Tools

Formerly known as Azure AI Services or Azure Cognitive Services is a unified collection of prebuilt AI capabilities within the Microsoft Foundry platform


Answer accepted by question author

Ramr-msft 17,836 Reputation points
2021-10-27T14:10:49.783+00:00

@Lalith Thanks, Please use the Opinion Mining feature of Sentiment Analysis, starting in version 3.1. Also known as Aspect-based Sentiment Analysis in Natural Language Processing (NLP), this feature provides more granular information about the opinions related to attributes of products or services in text.

To get Opinion Mining in your results, you must include the opinionMining=true flag in a request for sentiment analysis.
Here Sentiment Analysis might only report a positive sentiment in this case.

Was this answer helpful?

0 comments No comments

0 additional answers

Sort by: Most helpful

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.