PeaR.
adaptive Personality Recognition system
written by Fabio Celli,
university of Trento
current version: 4.0


HOW IT WORKS:
this system is based on adaptive personality recognition, a language-independent and fast algorithm for the extraction of personality types from text. Given as input a list of authors and texts, the system produces hypotheses on the personality of the authors.
Personality is formalized as a vector of 5 scores, corresponding to the traits of the big5 factor theory, the most widely accepted way to formalize personality in research:
- Extraversion (high=extrovert, low=shy)
- emotional Stability (high=calm, low=neurotic)
- Agreebleness (high=friendly, low=ugly)
- Conscientiousness (high=organized, low=careless)
- Openness to experience (high=insightful, low=unimaginative)
The system is language independent. It has been tested on English and Italian with performances ranging from 62% to 73% of correct predictions, depending on the the size and quality of the data.

INPUT FORMAT:
input files must be in a tab-separated format with many texts per authors, separeted with a |EOL| symbol (author[TAB]text|EOL|text), like this example.
upload your file (max 1Mb)

---

load a file

---