Social Networks can be an interesting window into the psychologies of friends, colleagues and total strangers, leaving a linear, analyzable record that can present a more cohesive picture of a person's personality – or desired personality – than a series of fragmented encounters in the physical world might ever do. TweetPsyche, created by Dan Zarella, takes advantage of “personality aggregation,” using psycholinguistic algorithms to create a psychological profile of someone based on their tweets. “I think the possibilities of a system like this are enormous, from matching like-minded users to identifying users that exhibit certain useful or desirable traits,” Zarella writes.
To see someone’s psychological profile, just type in their profile name and TweetPsyche does the analysis. Or you can use bookmarklet on a Twitter feed.
TweetPsyche compares a user’s tweets to a baseline of over 1.5 million random tweets to create a psychological profile based on two psycholinguistic methods, the Regressive Image Dictionary and the Linguistic Inquiry and Word Count.
The Regressive Image Dictionary (RID) follows a modified Freudian scheme, analyzing language for three categories: primordial (the unconscious) conceptual (logical and rational thought) and emotional. RID uses more than 3,000 words from 43 cognitive and emotional categories.
The second method is the Linguistic Inquiry and Word Count (LIWC), which “allows you to determine the rate at which the authors/speakers use positive or negative emotion words, self-references, big words, or words that refer to sex, eating, or religion.
The program was designed to analyze simply and quickly over 70 dimensions of language across hundreds of text samples in seconds.” Zarella combined these two methodologies with a Portor stemming algorithm and a Twitter analysis infrastructure.
“The service analyzes your last 1000 Tweets; as such, it works best on users who have posted more than 1000 updates,” writes Zarella.