Personality and emotion—A comprehensive analysis using contextual text embeddings
Personality and emotions have always been closely intertwined since humans evolved, adapting to these two forms. Emotions are indicative of a person’s personality, and vice versa. This paper aims to investigate the complex relationship between these two fundamental aspects of human behavior using th...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-12-01
|
| Series: | Natural Language Processing Journal |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2949719124000530 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850256732119367680 |
|---|---|
| author | Md. Ali Akber Tahira Ferdousi Rasel Ahmed Risha Asfara Raqeebir Rab Umme Zakia |
| author_facet | Md. Ali Akber Tahira Ferdousi Rasel Ahmed Risha Asfara Raqeebir Rab Umme Zakia |
| author_sort | Md. Ali Akber |
| collection | DOAJ |
| description | Personality and emotions have always been closely intertwined since humans evolved, adapting to these two forms. Emotions are indicative of a person’s personality, and vice versa. This paper aims to investigate the complex relationship between these two fundamental aspects of human behavior using the concepts of machine learning and statistical analysis. The objective is to automate the process of determining the relationship between personality traits of the MBTI (Myers-Briggs Type Indicator) and Ekman’s emotions based on the context of user-written social media posts using contextual embedding. A robust mechanism is employed, involving two main phases to figure out emotions from the social media posts. The first phase involves determining the cosine similarity scores between each MBTI personality trait and predefined emotions. The second phase introduces a cross-dataset learning approach where several machine learning models are trained on a dataset labeled with emotions to learn patterns of emotions found in the text. After training, these models utilize the patterns they learned to predict emotions in a targeted dataset. With an overall accuracy of 85.23%, the Support Vector Machine (SVM) is chosen as the most effective and high-performing model for emotion prediction tasks. We employed a vetting mechanism combining two approaches to improve accuracy, reliability, and trustworthiness for the final emotion prediction. Finally, using statistical quantification, this paper finds patterns that link each MBTI personality trait with Ekman emotions. It reveals that extroverts (E), sensing (S), and feeling (F) personality types are more likely to share joyful and surprising emotional posts, while individuals with extroversion (E), intuition (N), thinking (T), and perception (P) traits tend to express negative emotions such as anger and disgust. Conversely, introverts (I), intuitive (N), thinking (T), and judging (J) personalities are more inclined to share posts reflecting fear and sadness. This comprehensive study provides valuable insights on how individuals with different personality types typically express emotions on social media. |
| format | Article |
| id | doaj-art-d7ec666bf1f84e7598b0326564b516cf |
| institution | OA Journals |
| issn | 2949-7191 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Natural Language Processing Journal |
| spelling | doaj-art-d7ec666bf1f84e7598b0326564b516cf2025-08-20T01:56:34ZengElsevierNatural Language Processing Journal2949-71912024-12-01910010510.1016/j.nlp.2024.100105Personality and emotion—A comprehensive analysis using contextual text embeddingsMd. Ali Akber0Tahira Ferdousi1Rasel Ahmed2Risha Asfara3Raqeebir Rab4Umme Zakia5Ahsanullah University of Science and Technology, 141 and 142, Love Road, Tejgaon, 1208, Dhaka, Bangladesh; Corresponding author.Ahsanullah University of Science and Technology, 141 and 142, Love Road, Tejgaon, 1208, Dhaka, BangladeshAhsanullah University of Science and Technology, 141 and 142, Love Road, Tejgaon, 1208, Dhaka, BangladeshAhsanullah University of Science and Technology, 141 and 142, Love Road, Tejgaon, 1208, Dhaka, BangladeshAhsanullah University of Science and Technology, 141 and 142, Love Road, Tejgaon, 1208, Dhaka, BangladeshNew York Institute of Technology, Vancouver Campus, Vancouver, CanadaPersonality and emotions have always been closely intertwined since humans evolved, adapting to these two forms. Emotions are indicative of a person’s personality, and vice versa. This paper aims to investigate the complex relationship between these two fundamental aspects of human behavior using the concepts of machine learning and statistical analysis. The objective is to automate the process of determining the relationship between personality traits of the MBTI (Myers-Briggs Type Indicator) and Ekman’s emotions based on the context of user-written social media posts using contextual embedding. A robust mechanism is employed, involving two main phases to figure out emotions from the social media posts. The first phase involves determining the cosine similarity scores between each MBTI personality trait and predefined emotions. The second phase introduces a cross-dataset learning approach where several machine learning models are trained on a dataset labeled with emotions to learn patterns of emotions found in the text. After training, these models utilize the patterns they learned to predict emotions in a targeted dataset. With an overall accuracy of 85.23%, the Support Vector Machine (SVM) is chosen as the most effective and high-performing model for emotion prediction tasks. We employed a vetting mechanism combining two approaches to improve accuracy, reliability, and trustworthiness for the final emotion prediction. Finally, using statistical quantification, this paper finds patterns that link each MBTI personality trait with Ekman emotions. It reveals that extroverts (E), sensing (S), and feeling (F) personality types are more likely to share joyful and surprising emotional posts, while individuals with extroversion (E), intuition (N), thinking (T), and perception (P) traits tend to express negative emotions such as anger and disgust. Conversely, introverts (I), intuitive (N), thinking (T), and judging (J) personalities are more inclined to share posts reflecting fear and sadness. This comprehensive study provides valuable insights on how individuals with different personality types typically express emotions on social media.http://www.sciencedirect.com/science/article/pii/S2949719124000530MBTI (myers-briggs type indicator)Ekman emotionsCosine similaritySupport vector machine (SVM)Personality traitEmotion detection |
| spellingShingle | Md. Ali Akber Tahira Ferdousi Rasel Ahmed Risha Asfara Raqeebir Rab Umme Zakia Personality and emotion—A comprehensive analysis using contextual text embeddings Natural Language Processing Journal MBTI (myers-briggs type indicator) Ekman emotions Cosine similarity Support vector machine (SVM) Personality trait Emotion detection |
| title | Personality and emotion—A comprehensive analysis using contextual text embeddings |
| title_full | Personality and emotion—A comprehensive analysis using contextual text embeddings |
| title_fullStr | Personality and emotion—A comprehensive analysis using contextual text embeddings |
| title_full_unstemmed | Personality and emotion—A comprehensive analysis using contextual text embeddings |
| title_short | Personality and emotion—A comprehensive analysis using contextual text embeddings |
| title_sort | personality and emotion a comprehensive analysis using contextual text embeddings |
| topic | MBTI (myers-briggs type indicator) Ekman emotions Cosine similarity Support vector machine (SVM) Personality trait Emotion detection |
| url | http://www.sciencedirect.com/science/article/pii/S2949719124000530 |
| work_keys_str_mv | AT mdaliakber personalityandemotionacomprehensiveanalysisusingcontextualtextembeddings AT tahiraferdousi personalityandemotionacomprehensiveanalysisusingcontextualtextembeddings AT raselahmed personalityandemotionacomprehensiveanalysisusingcontextualtextembeddings AT rishaasfara personalityandemotionacomprehensiveanalysisusingcontextualtextembeddings AT raqeebirrab personalityandemotionacomprehensiveanalysisusingcontextualtextembeddings AT ummezakia personalityandemotionacomprehensiveanalysisusingcontextualtextembeddings |