Cross-Cultural Personality Prediction based on Twitter Data

Abstract

Abstract:Social networking platforms such as Facebook, Twitter, YouTube, and Instagram, which generate a vast amount of data every second, emerged dramatically within the last ten years. This huge rich data provides crucial information about social interactions and human behaviour. Therefore, it is possible to identify the personality traits of a person by extracting and analysing relevant information from the social media. Recently, researchers demonstrated that personality prediction can be performed by using historical textual features and user profiles. While cross-cultural personality research is considered as a powerful tool to observe the differences in personality psychology, machine learning researchers analysing social media data did not perform research on the development of cross-cultural personality prediction models yet. In this paper, we propose a new research topic called cross-cultural personality prediction and discuss how these kinds of models can be built in practice. In the joint research project, we investigate whether there is a cultural or language difference in personality between people in two countries namely, South Korea and Turkey in Twitter. It will be particularly interesting because both of the countries are in Asia, but located in both ends of Asia and we'll investigate whether there is any meaningful personality difference in two countries or not. We argue that this kind of research will impact not only personality services such as IBM Personality Insight, but also cross-cultural personality psychology research area in the near future.

Modeling Personality vs. Modeling Personalidad

“…In our models, however, Call features did not turn out to be predictive for these sociability-relevant traits. This might be due to a recent and considerable shift to Internet based call and messenger services such as FaceTime, WhatsApp, Viber and similar mobile apps 6 .…”

“…The past decade has seen an increase in the literature on automatic personality recognition, especially in modelling personality traits from data generated by diverse digital and real-world behaviors [56]. For instance, a large body of research suggests that data generated from the use of social media websites such as Facebook, Twitter, and Blogging Sites can be used to accurately infer personality traits [6,20,28,43,52]. The value of digitally derived measurements for personality assessment was exemplified in studies showing that Facebook Likes could be used to achieve higher accuracy, compared to personality scores provided by human raters [38,59].…”

Kosinüs Benzerliğine Dayalı Çapraz-proje Hata Tahmini

“…Yöntem ilgili deneysel veri setlerinde AUC açısından tahmin sonuçlarını kayda değer bir biçimde arttırmıştır. Benzer bir çalışma Catal ve arkadaşları [21] tarafından özellik seçimini sağlayan yeni bir algoritma önerilerek yapılmıştır. Ancak kullanılan veri setindeki gürültü oranı ve yanlış etiketli örnekler nedeniyle sonuçlar yanlış hesaplanmış veya yorumlanmış olabilir.…”

Personality Is Revealed During Weekends: Towards Data Minimisation for Smartphone Based Personality Classification

“…Psychologists showed that personality traits capture stable individual characteristics that explain and predict behavioural patterns [6]. Interestingly, personality traits can also predict patterns of technology use, such as behaviours in social media [3,9], blogs [14], games [29], phone use [4,15,24] and even how users choose app permission settings [21]. Therefore, personality is considered to be relevant to a number of computing areas, among which Human Computer Interaction (HCI) can particularly benefit from understanding users' personality, by making informed decisions about their needs and preferences.…”

“…When it comes to user modelling, app designers typically avoid using this method as completing questionnaires can be cumbersome for users and can consequently drive them away from the app. For this reason, automatic prediction of personality has attracted the attention of many scholars and practitioners who relied on data collected from Twitter [3], Instagram [9], blogs [14], and smartphone use [4,15,24]. Most of these approaches relied on collecting data from several weeks [24], months [17] and even years [15,16], in order to accurately infer personality.…”