We are introducing a
new demo for
IBM Watson Personality Insights, a
Watson Developer Cloud service that lets you understand different individual’s personality characteristics, solely based on text analysis. This new demo adds features to maximize understanding of personality characteristics and to improve the overall user experience, focusing on the relation between personality traits and behaviors that a person is more or less likely to express.
Read: IBM Watson Personality Insights General Availability Announcement
In addition to the options of writing your own text to be analyzed or copying from a different source, we have extended the text samples available. Since Twitter is one of the most popular media sources that users have been using to analyze and draw conclusions from audiences, we have added three celebrities who are frequent Twitter users as samples, on which you can perform the personality analysis.
Just select a sample and hit Analyze! Together with the personality summary, which provides an agile way to understand the outcome of the analysis, you will also visualize feedback about how strong the results are, a list of the behaviors that the personality is likely or unlikely to manifest, and the lists of traits and facets of the analyzed personality separated by category.
The analysis strength is based on the number of words needed to get statistically significant results. While we will allow you to invoke the service with as little as 100 words, more words are needed to get reliable results. There are four levels of strength: Weak (100~1500 words), Decent (1500~3500 words), Strong (3500~6000 words) and Very Strong (6000+ words).
The behaviors shown are based on
studies by researchers which reveal different correlations between personality traits and behaviors in certain industries or domains. Their findings include facts such as: Spending habits are related to the emotional range trait, risk profiles to extraversion or introversion, and healthy decisions are related to conscientiousness, to name a few.
The full personality scoreboard features all the trait values produced by the service analysis, allowing a granular view of the personality that will let you explore each individual aspect of the personality, with the possibility of a more complex analysis. These traits are grouped into the three kind of personality insights provided by the service: Personality, Consumer Needs and Values. Each trait value is a percentile score which compares the individual to a broader population.
Any questions about the meaning and implication of a behavior or trait? Mouse over the word and a tooltip will enlighten you. Give the
new demo a try!
For further information see the
usage guidance for the Personality Insights service.
