The sentiment visualization feature can be leveraged to gain an understanding of the tone of the social media content found during analysis. Sentiment is grouped by the topics that were specified when setting up the project. Relative sentiment shows the dispersion of constructive, undesirable, unbiased, and undecided sentiments. Sentiment terms are phrases that measure the feel of a mention. Sentiment shows whether a mention is good or bad. A mention is branded as undecided when it has the same number of good and bad sentiment phrases. A mention is branded as neutral when no sentiment terms are detected in it.

The sentiment trend chart illustrates the number of mentions by day, spanning the time range of our project, as shown in the following screenshot:

In our project, the visualization indicates to us that positive mentions are more common for social content mentioning Amazon customer service. This may be a bit clearer if we filter the visualization (right-click and select the Exclude button) by removing the neutral responses, as shown in the following screenshot:

Now the visualization is looking better: