Muck Rack automatically analyzes articles that are added to Coverage Reports and assigns a sentiment of positive, negative, or neutral to each article. The sentiment assigned to an article is based on scoring applied to the words in the headline and content of the article. Scores applied to the content of an article are compiled and analyzed for special cases, like when an author has used negation, superlatives, or idioms.
Editing an Article's Sentiment Score
If Muck Rack’s assigned sentiment score doesn’t match your own analysis of the article, you can easily edit the sentiment assigned to be reflected in your Coverage Report for an individual article or in bulk.
Select the Articles tab in your Coverage Report.
Select the checkbox(es) icon to the left of the article(s) you’d like to edit.
Select Edit.
Choose a new sentiment score from the dropdown.
Editing the sentiment of an article in one Coverage Report will not impact the sentiment of the article in your other Coverage Reports. That way, if an article mentions different clients, campaigns, or key people you can adjust the sentiment for the article in each Coverage Report as desired.
Show or Hide Sentiment Within Your Coverage Reports
Don’t want to see Sentiment in your Coverage Reports and exports? You can easily show or hide the sentiment analysis in your Coverage Report by following the steps below.
Navigate to the Articles tab of your Coverage Report.
Select Edit Columns at the top-left of the page.
From the list of metrics, select ARTICLE DATA and toggle Sentiment on or off.
Once this is toggled ON, Sentiment will appear in:
the Articles tab of your Coverage Report
Presentations made with that report
your Excel and PDF exports for that report
Suggested Workflow
We suggest the following workflow for sentiment tracking and reporting in Muck Rack:
Ensure articles are automatically added to your Coverage Reports.
Review those articles and make any necessary sentiment adjustments.
Create Dashboard widgets that showcase the sentiment for a given report.
Create a Presentation to showcase this impact to stakeholders.
What technology does Muck Rack use to determine Sentiment?
Muck Rack uses Natural Language Toolkit, an open-source natural language processing library, to calculate article sentiment. When an article is added to a Coverage Report, we parse the text of that article and assign a score to each word in the article based on its polarity and perceived strength ("terrible" is stronger than "bad"). The scores of each word are then combined to derive the overall score for the article. The overall score is a numeral to which we then assign the sentiment label of positive, negative, or neutral.