Finally, an objective resource that generates independent scoring about difficult, divisive subjects. The Civility Scorecard is designed to encourage civil speech that opens the door to work more productively with a common understanding of how people are communicating, without the influence of our own subjective bias.
An objective solution to measure the relative level of civility of a speech based on machine learning and artificial intelligence.
Paste the speech into the text box, provide the speech information and submit.
A few seconds later you will receive the civility score and a detailed write up.
An indication of where the speech falls on the civility scale is provided.
Nothing brain taxing, we do all the hard work and heavy lifting to provide you with insight.
A detailed write up on what is being “said” in the speech, but ignoring the content words.
Understand changes over time, see how the personality of the speaker changes.
Using psychology, linguistics as well as machine learning we are able to look at the function words used to unearth the message behind the speech.
“Content” words speak to the topic of discussion and account for only 40% of what is being said. “Function” words account for 60% of the content and provide deeper insight.
The Civility Score is a ranking between 10 and 90, as no speech is completely civil or uncivil, and measures how civil a speech is.
Function words are the shortest and most common words that we use. They include pronouns, articles, prepositions, conjunctions, negations, auxiliary verbs, and a small group of common adverbs. Although there are only about 500 function words, they account for over 60% of all the words you say, hear, or read.
They are also unconscious words, people do not think about these words when using them, and therefore they provide deeper insight into the unconscious of what is being said.
With an understanding of what speeches in history have been civil and those that have not, we were able to apply machine learning, psychology and linguistics to allocate scores to speeches and place them on a continuum ranking between 10 and 90. We use this dataset to score new speeches that are loaded and use the new data provided to help us further refine the scores into the future.
We have used a variety of speeches since the transcription of speeches have been recorded. Ranging from speeches by Martin Luther King, Nelson Mandela, Joe Biden, Donald Trump, Barak Obama, various television commentators including Fox and CNN, The Black Panthers, Hitler, Churchill, Roosevelt and JFK.
Some of the speeches that have been scored:
The presidential acceptance at 61%
The presidential acceptance at 58%
A very uncivil speech scoring 34%