The Sentiment Analysis Recipe and Dataset

The Sentiment Analysis template creates a recipe that performs sentiment analysis on the source object field that you selected in the configuration wizard. This recipe outputs a dataset with the analysis results and any additional selected fields.

The recipe applies the detect sentiment transformation to the selected text field and assigns a value of Positive, Neutral, or Negative. For more details, search for and open Detect Sentiment Transformation.

Note Detect Sentiment works on a single dimension column and supports English only. It processes non-English text as English and ignores images, including emoji. If the input value is null or empty string (‘’), the transformation assigns the sentiment as Neutral or with a value of 2.5, depending on the output type. Applying Detect Sentiment on text such as IDs, nouns, addresses, or alphanumeric values can return unexpected results.

The output dataset, SentimentData, contains the sentiment analysis result in the sentiment column, labeled with the field name appended with Sentiment. All additional fields are also present in the output dataset.

You can customize the templated recipe in the recipe editor. If you make changes, it’s recommended that you save it as a new recipe.