Donald Trump has become the 45th President of the United States ever since he grew up as an outsider of American politics.

He is one of the politicians to utilize social media very well, out of which Twitter has been the best gear to deliver his thinking to the public.

By using the feature of ‘NetMiner Semantic Network Analysis’, we analyzed Trump’s Twitter messages posted after he won the presidential campaign, about what he has tried to deliver to the public and what he has emphasized most.

At first, we collected approximately 1200 tweets posted from Nov 2016 to Jun 2017 in Donald Trump's account (@realDonaldTrump).

His monthly tweets posted are as follows:

We found that Mr. Trump has posted about 150 tweets per month and he posted the most at the time of inauguration.

So, what are the most frequent words in his tweets?

As a result of extracting words from tweets by using NetMiner, those are 'United States, Job, White House, news, election, HealthCare, GREAT, AGAIN'.

For your reference, we extracted only nouns through eliminating unnecessary words by using ‘Filter & Dictionary’, and out of multiple synonyms we chose one representative word by using ‘Thesaurus’. 

We found that the word of ‘United States’ is the most and ‘country’, ‘job’ are the next that are frequently mentioned, which represents that he was strongly supported by the white, working-class voters under the slogans of ‘America First’.

Through the words of ‘news ’, ‘media’ and ‘Fake’ appearing in a sizable proportion, we can also guess his thoughts on the mainstream media.

And we create a link between words by calculating how closely two words appear in a sentence.

In the following map, the size of the dot reflects the frequency of the word, and each shape of words different from a circle are those classified as a proper noun such as geographical names, organization names, personal names.

NetMiner recognizes and classifies the names of geography, organization, person by itself and returns them as the property of the words.

You can see the word ‘HealthCare’ is centered among the words of ‘plan’, ‘insurance’ and ‘bill’. Also ‘news’, ‘fake’, ‘media(medium)’, ‘election’, ‘Russia’ are closely located in the word network.

And another interesting thing to look at is ‘job’ is connected with many other words by line. It means ‘job’ is frequently mentioned together with those words.

You can also see ‘MAKE’, ‘AMERICA’, ‘GREAT’, and ‘AGAIN’ are condensed together. This reflects on Mr. Trump’s focus, “Make America Great Again (MAGA)”.

And next, let’s figure out what the word ‘job’ is associated with and what are the tweets he mentioned about 'Job'. At first, ‘Job’ is related to ‘United States(Country, America)', ‘business’, ‘billion’, ‘dollar’, ‘plant’, ‘progress’ and ‘wealth’.  

You can see Mr. Trump's idea that 'jobs' create America's wealth.

Using ‘Words in a sentence’ of NetMiner, we can also find out what sentences ‘job’ appears in.

Through it, we recognize that he emphasized 'job' so much enough to mention it alone in some cases without any other words.

This means jobs are what his supporters needed most and therefore he has focused intensively on it by using the word ‘job’ frequently in his tweets.

Finally, through topic modelling analysis, we identified the potential topics embedded in Trump's tweets.

The following diagram depicts the sub-topics in Trump’s Tweets and their composing keywords.

In conclusion, Trump has focused through 1200 posts mostly regarding to ‘Job’, ‘News’, ‘HealthCare’ and ‘Russia Scandal’ since he was elected president.

The interesting thing is that China is related to the subject of Job.

Does Mr. Trump regard China as one of the major reasons why U.S. manufacturing industry has declined and workers lost their jobs? If so, how will US-China relation change while he is in office?  

As you can see so far, by using text analytics, you can easily summarize large volume of text and find out its subject quickly, even without reading them through.

At this time by using NetMiner, we see ‘How President Trump Thinks’. We will come back to you next time with more interesting subject.

Thank you very much.

* All the analysis output are produced by using the ‘Semantic Network Analysis’ of NetMiner.

* Download NetMiner evaluation kit