What is the semantic rating?
The semantic rating identifies emotional keywords patients use in their open-text survey responses. Emotional keywords such as “happy” or “sad” are assigned a value between -100 and 100.
What is the sentiment rating?
Sentiment analysis provides context and clarity with a quick look at the overall tone of the whole comment, rather than just individual keywords. The sentiment rating uses an algorithm to give you a better sense of your comments containing negation words such as ‘not’ in conjunction with negative keywords such as ‘afraid’ (for example, ‘The way they talked to me made me not feel so afraid’). A comment might have a negative semantic score but a positive sentiment score.
The bar chart at the top of the comments list gives you an overview of the semantic ranking for each of your selections.
To quickly identify the most positive or negative comments, click on any of the Semantic or Sentiment column headers to sort the information. Each comment offers information about the provider, the date it was submitted, and the location. You may choose to view comments from a specific timeframe to narrow down the results.
How is comment sentiment measured?
Our sentiment analysis uses natural language processing (NLP) and machine learning algorithm provided by Amazon Web Services. This algorithm is constantly learning and adapting to give the most accurate sentiment analysis possible.
Why are some comments rated ‘Negative’ when they actually aren’t?
Depending on the individual words used in the comment, the algorithm may misidentify a negative tone in a positive comment. This is a less common occurrence that will improve over time as the algorithm analyzes an increased amount of comments.
Sentiment ratings do not affect your survey composite score. These ratings are intended to help you more quickly analyze the overall tonal insights of your patient comments, but are not calculated in any survey scoring.