Social Sentiment Analysis in Microsoft Social Engagement calculates the sentiment value of social posts utilizing natural language processing and machine learning techniques.
In Social Engagement Analysis, go to Analytics > Sentiment to learn more about Social Sentiment Analysis across posts in your data set.
Sentiment value is the negative, positive, neutral, or unknown sentiment for a post. Once in a while, the algorithm identifies positive and negative parts of a sentence and rates the post as neutral, because the amount of positive or negative text cancel each other out.
You can edit and confirm Social Engagement sentiment values for individual posts to benefit from adaptive learning for your organization’s sentiment algorithm. An unknown Social Sentiment Analysis value indicates that the language isn’t supported by the sentiment algorithm.
Sentiment index is going to calculate from the sentiment value of posts and normalized to a value between -10 and 10. All your active filters and parameters are considered to characterize the data set that the sentiment index calculates.
- A sentiment index of 10 implies that there are no negative posts in your data set.
- Sentiment index of 0 implies that there are equal amounts of positive and negative posts in your data set.
- A sentiment index of -10 implies that there are no positive posts in your data set.
The typical view of this widget shows the top five authors and sources, based on the volume of posts and trend indicator. Click on the Full view button to extend the widget and find more details such as reach, source, and location about the 100 most-active authors and their posts.
To include a filter for the multiple authors at once, select the check boxes on the left side of the list for all authors that you want to include. Then select INCLUDE in the list header. To expel an author from the authors filter, select the check boxes on the left side of the list for all authors that you want to remove from the filter. Then select EXCLUDE in the list header.
Full view additionally has a Delete button that you can use to delete a selected author and the author’s posts. You should have a Power Analyst or Administrator user role to delete an author. When you remove an author, none of the authors posts will be available in the solutions database meaning they are permanently deleted.
- Location Insights
Posts that have location information display on a map showing where the posts are coming from. You can also include an activity map to see new posts in real-time, with additional functionality. Dynamic widgets shows only if the posts having location information that are available in the selected data set.
- Negative Phrases
Lists regularly mentioned negative phrases, based on the posts in your current data set. The larger that a phrase shows up, the more posts contain this phrase.
- Positive Phrases
Lists regularly mentioned positive phrases, based on the posts in your current data set. The larger a phrase shows up, the more posts contain this phrase.
Sentiment for Social Engagement Analysis
Visualize the index across all posts that have a sentiment value in the selected data set. It additionally demonstrates the change in sentiment index compared to the last similar time frame and trend indicator. You can physically change the sentiment values if you find that a post’s sentiment is analyzed incorrectly.
- Sentiment Coverage
Demonstrates the relative distribution of system-rated and manually edited sentiment values.
- Sentiment History
It shows the number of posts with sentiment index, the sentiment value, and the average sentiment value for a selected time frame.
- Sentiment by Source
List out the sentiment index for the top five sources and the trend indicators.
- Top Critics
List out the authors who published the most posts with a negative sentiment value.
- Top Fans
List out the authors who published the most posts with a positive sentiment value.
Locus IT is a Microsoft Dynamics partner and provides Social Sentiment Analysis services for Microsoft Social Engagement along with Sentiment Analysis training, Sentiment Analysis implementation and Sentiment Analysis support. For more information please contact us.