| A | |
|---|---|
1 | Methodology To create this table we made the list of 50 notable CEOs and business leaders, plus Donald Trump and Joe Biden. Then we collected multiple transcripts where the leader in question speaks. We purified the transcripts by removing all text that is not the leader’s speech. We have ensured the minimum length of the transcripts 10,000 for each leader. After that we took 10,000 words sample for each leader and counted the number of unique words in this sample. Then using the pre-labelled emotion association lexicon (http://saifmohammad.com/WebPages/NRC-Emotion-Lexicon.htm) we counted the percentage of the words that are positive / negative as well as are associated with one of eight emotions (this lexicon consists of only 14,182 words; in average about 24% of the words from our transcripts are presented in the given lexicon; so the percentage representation of emotionally associated words to the total number of words found in the lexicon best reflects the emotional colour of speech). To identify top words of Trump and Biden we counted average count of words among all 52 leaders, left only 2,000 more frequent words and then calculated the ratio of words count in Trump / Biden transcripts to the average count. Words with the biggest ratio are presented on these two tabs. Original source: https://businessfinancing.co.uk/leader-vocabulary/ |