Personality prediction model for social media using machine learning Technique. (May 2022)

• Record Type: Journal Article
• Authors: Kamalesh, Murari Devakannan; B, Bharathi
• Is Part Of: Computers & electrical engineering. Volume 100(2022)
• Publication Date: 2022-05
• Subjects: Personality prediction -- Big five personality traits -- Binary-Partitioning Transformer (BPT) -- Term Frequency & Inverse Gravity Moment (TF-IGM); Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2022.107852
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights: Exhibiting the relationship between users' personalities with their behavior in the interactions of social media platforms. Predict the personality traits to explore in the social media platform using machine learning approach. The current work comprises five stages: data collection, pre-processing, Feature extraction, and selection, using machine learning algorithm. The proposed feature extraction method using Binary-Partitioning Transformer(BPT) with the TF-IGM approach better predicts personality traits. Abstract: Predicting human behavior and personality from the social media applications like Facebook, Twitter and Instagram is achieving tremendous attention among researchers. Statistical information about the human thoughts expressed via status on social media is essential assets for research in predicting various human behaviour and personality. The current work mainly focuses on guessing user personality based on big five personality traits. An intelligent Sentence analysis model is built to extract personality features. In this article, a new Binary-Partitioning Transformer (BPT) with Term Frequency & Inverse Gravity Moment (TF-IGM) is proposed that identifies relationships among feature sets and traits from datasets. The proposed work outperforms the all feature extraction average baseline set on multiple social datasets. A maximum F1-score of 0.762 and accuracy of 78.34% on the Facebook dataset; 0.783 and 79.67%; on the Twitter dataset, 0.821; 86.84% on Instagram dataset is achieved. Graphical abstract: Image, graphical abstract