TwitPersonality: Computing Personality Traits from Tweets Using Word Embeddings and Supervised Learning

Abstract

Abstract:We are what we do, like, and say. Numerous research efforts have been pushed towards the automatic assessment of personality dimensions relying on a set of information gathered from social media platforms such as list of friends, interests of musics and movies, endorsements and likes an individual has ever performed. Turning this information into signals and giving them as inputs to supervised learning approaches has resulted in being particularly effective and accurate in computing personality traits and types. Despite the demonstrated accuracy of these approaches, the sheer amount of information needed to put in place such a methodology and access restrictions make them unfeasible to be used in a real usage scenario. In this paper, we propose a supervised learning approach to compute personality traits by only relying on what an individual tweets about publicly. The approach segments tweets in tokens, then it learns word vector representations as embeddings that are then used to feed a supervised learner classifier. We demonstrate the effectiveness of the approach by measuring the mean squared error of the learned model using an international benchmark of Facebook status updates. We also test the transfer learning predictive power of this model with an in-house built benchmark created by twenty four panelists who performed a state-of-the-art psychological survey and we observe a good conversion of the model while analyzing their Twitter posts towards the personality traits extracted from the survey.

A large-scale, in-depth analysis of developers’ personalities in the Apache ecosystem

“…To date, there has been limited but growing amount of work on the automatic detection of personality traits from conversation transcripts and written text [44]. Thanks to the advancements in artificial intelligence (AI) and the widespread diffusion of social media contents, researchers have explored methods for the automatic recognition of various types of pragmatic variation in text and conversations, both short-lived, such as emotion, sentiment, and opinions [45][46][47], and more long-term, such as personality [48,49].…”

“…Finally, Carducci et al [48] developed TwitPersonality, 7 a personality detection model that uses word vector representations of tweets fed to SVMs. The Twitter histories of 24 volunteers were retrieved along with their Big Five personality traits, measured using the BFI questionnaire.…”

Multilingual Transformer-Based Personality Traits Estimation

“…In [8], Carducci et al created a Support Vector Machine (SVM) to perform a supervised regression with the myPersonality dataset. They developed and fine tuned a SVM with 300-dimensional word embeddings as feature vector and each personality trait score as target.…”

“…As shown in Section 6, we reduce significantly the mean squared error in trait prediction and we also create a smoother data distribution of the predicted scores. In fact, the scores at the tails of the data distribution are detected more correctly than previous models [6,8,9]. Our stacked neural network receives as input the 768-dimensional CLS token from the encoding phase in our pipeline, as illustrated in Figure 2.…”

MBTI Based Personality Prediction o f a User Based o n Their Writing o n Social Media

“…The accuracy rates are the highest for Experience, Openness, Neuroticism and Extraversion (over 84%), whereas for Agreeableness and Conscientiousness, the accuracy is 77%. Giulio Carducci et al [9] modelled a paper that determines the various personality traits of an individual based on their tweets, this was done through a supervised learning method. The methodology that they followed first breaks the tweets down to numerous tokens through which it then learns the representation of various words as embeddings.…”