Browse Prior Art Database

Predicting Personality Traits based on Pictures in User's Social Media

IP.com Disclosure Number: IPCOM000249259D
Publication Date: 2017-Feb-15
Document File: 2 page(s) / 62K

Publishing Venue

The IP.com Prior Art Database

Abstract

This article proposes a method to predict personality traits based on text and pictures from user’s social media by extracting pictures features including main description, background, social hotspot, timeline of the pictures, and emotions of the user. And then we build a predictive personality model based on text and picture features.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 59% of the total text.

1

Predicting Personality Traits based on Pictures in User's Social Media

Personality analysis requires quality data especially text related to psychology. However, in some applications, data source is poor. For example, in social media, some people may use many pictures with a little description. Personality result of existing method may be inaccurate.

Existing works identify personality based on demographic features, e.g. age, gender, or predict personality using psycholinguistics models. And directly converting pictures to description may not work. One prior art analyzed personality through social media profile picture choice; some recently proposed methods also consider emotion, sentiment, or speech. But they do not analyze pictures in user’s social media.

Pictures shared in social media may be closely related to personality, for example  Picture content may related to personality

– People with high extraversion may be more willing to show pictures related to himself or herself or children or family  Background of the pictures

– Cheerful, sunny VS Deserted, dark  Social hotspot related to the pictures

– Many pictures are related to social hotspots, may be high neuroticism  Timeline of the pictures

– Compared with other for the same topic, publish later than others  Emotion of the owner

– Extract the owner’s face from pictures in his/her social media – Extract emotion from these face picture – No or very litter owner’s face

 Text or description is the same as extract...