IBM’s Watson User Modeling Service “uses linguistic analytics to extract a spectrum of cognitive and social characteristics from the text data that a person generates through text messages, tweets, posts, and more.”

Input a chunk of text, and get back your Big Five and more.

Graphic source:
Try it yourself!

It looks like it’s designed for (creepy, intrusive) marketing purposes, but the demo shows an analysis of a fictional character.  Hey, why not?

How did it do with my main character, Molly Barda?

Molly earned her Ph.D. from one of the top Literature and Creative Writing programs in the country. After a year of fruitless job-hunting (no one warned her about that!), she finally found a position teaching business communication in the College of Commerce at remote Mahina State University. She is determined to make a success of her new life and to bloom where fate, and the academic job market, have planted her.

I fed a chunk of Molly’s dialog into the site. According to the Watson User Modeling Service, Molly ranks above the 75th percentile in these traits:

Self-expression 88%
Fiery 86%
Structure 86%
Practicality 85%
Authority-challenging 81%
Orderliness 80%
Cautiousness 79%
Emotional range 77%
Susceptible to stress 77%
Need for Closeness 76%

She ranks low, below the 25th  percentile, in:

Excitement-seeking 2%
Ideal 2%
Gregariousness 8%
Liberty 8%
Hedonism 10%
Self-discipline 14%
Conservation 15%
Dutifulness 20%
Extraversion 20%
Outgoing 23%

This makes sense; Molly is supposed to be a neurotic INTJ personality type, emphasis on the I(ntrovert). Well done, Watson, and well done me, for drawing the character so well. But wait! What if that just happens to be the way I write? Would all of my characters come out with the same profile?

Let’s try a different character: Davison Gonsalves, the most obnoxious and unrepentant cheater in Molly’s Intro to Business Management class.

He is above the 75th percentile in:

Agreeableness 100% (He is a world-class suckup)
Uncompromising 100%
Outgoing 98% (this is very different from the introverted Molly)
Trust 95%
Extraversion 94%
Assertiveness 92%
Cheerfulness 91%
Altruism 91%
Orderliness 88%
Sympathy 87%
Modesty 86% (this one doesn’t seem right. Watson, explain?)
Gregariousness 84%
Fiery 76%

He is below the 25th percentile in:

Imagination 1%
Authority-challenging 5%
Achievement Striving 9%
Adventurousness 10%
Intellect 13%
Artistic 14%
Excitement-seeking 14%
Conscientiousness 18%
Activity 21%
Self-discipline 25%

He’s low in intellect, conscientiousness, imagination, and self-discipline. That sounds right.

Interesting that although he’s badly behaved, Davison is low in authority challenging, while Molly is high in authority challenging. I’d say that Davison wants to climb to the top of the ladder as quickly as possible, taking shortcuts wherever he can, while Molly would probably question the assumption that success needs to look like a ladder in the first place.

I think this could be a great tool for helping writers to keep characters’ dialog distinct, and to reflect on what kind of personality traits are coming through when each character speaks.


Blog  | Facebook  | GoodReads | LinkedIn | Twitter | Mailing List

2 thoughts on “Personality revealed through text–an interactive tool to analyze your characters (or yourself)

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.