Unearth the power of declared data
Insights That Work
Brand & Retailer tickets for all IIeX events now start at just $99! Get or give one today!

Sentiment Analysis is Simple (a Trump’s Tweets Example)

Sentiment analysis allows you to quickly gauge the mood of the responses in your data.

By Chris Facer

Social media provides a sea of information, and it can be hard to know what to do with it all. When people post their ideas and opinions online, we get messy, unstructured text. Whether its comments, Tweets, or reviews, it is costly to read them all. Sentiment analysis allows you to quickly gauge the mood of the responses in your data. This article takes a brief look at what sentiment analysis is, and applies some simple sentiment analysis to Donald Trump’s tweets.

What is sentiment analysis?

At its simplest, sentiment analysis quantifies the mood of a tweet or comment by counting the number of positive and negative words. By subtracting the negative from the positive, the sentiment score is generated. For example, this comment generates an overall sentiment score of 2, for having two positive words:
Positive sentiment analysis example


You can push this simple approach a bit further by looking for negations, or words which reverse the sentiment in a section of the text:

Negative sentiment analysis example

The presence of the word don’t before like produces a negative score rather than a positive one, giving an overall sentiment score of -2.

The process of reducing an opinion to a number is bound to have a level of error. For example, sentiment analysis struggles with sarcasm. But when the alternative is trawling through thousands of comments, the trade-off becomes easy to make.  A little sentiment analysis can get you a long way. This is especially true when you compare the sentiment scores with other data that accompanies the text.

Trump’s tweets

During the election campaign of 2016, much discussion revolved around who was sending out Donald Trump’s Tweets. A number of articles described how the tone of Trump’s tweets is more positive when they come from an iPhone device, than when they come from an Android. The hypothesis is that Trump tweets from an Android device, and that he employs social media assistants who tweet from an iPhone. But how do you work that out?

You add the sentiment scores to a data set, and then compare the sentiment scores for the different devices. You can try this example out for yourself in Displayr.

In a data set containing 1,512 tweets from @realDonaldTrump sent during the primaries, there is a small but positive average sentiment score of 0.3, with scores ranging from -5 to 6. This means that the average tweet has slightly more positive language than negative. The magnitude of the scores is small as the length of a tweet is restricted.

The power of sentiment arises when considering other variables in the data. Think of the now-famous example of the Trump sentiment gap between Android and iPhone. The mean sentiment score of Tweets from Android, 0.1, is significantly lower than the overall average of 0.3:

If these mean scores don’t sway you, then you may find the shape of the distribution more convincing:

The iPhone has a greater proportion of neutral (0) and slightly-positive (1) tweets. The Android has fewer such tweets, and a greater proportion of tweets with a negative score.


The data from Twitter includes the number of times each tweet has been Favorited. This is used as a proxy for engagement. For this data set, the average is around 19,000. By considering how the average number of favorites varies with the sentiment, we discover another interesting pattern.

Those tweets which have a negative sentiment (scoring -2 or fewer) garner a significantly higher number of favorites on average. It would seem that Trump’s followers are noticeably more engaged by negative content.

A little sentiment analysis can reveal patterns in the data which would be difficult to gain by reading through the sea of content.

You can analyze Donald Trump’s tweets yourself by clicking here


Emojis in this article come from the open-source Thanks to David Robinson for his blog post which inspired my recent thinking on sentiment analysis and text analysis.

This post was originally posted here

Please share...

Join the conversation