On Text-based Personality Computing: Challenges and Future Directions

Fang, Qixiang; Giachanou, Anastasia; Bagheri, Ayoub; Boeschoten, Laura; Kesteren, Erik–Jan van; Kamalabad, Mahdi Shafiee; Oberski, Daniel L.

doi:10.18653/v1/2023.findings-acl.691

Cited by 1 publication

(2 citation statements)

References 51 publications

(61 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Personality detection is a complex and difficult task, since the personality traits are latent theoretical variables that cannot be directly or objectively observed (Fang et al 2023). Previous research has demonstrated a strong link between personality traits and a person's sentiments, words, and opinions (Kishima et al 2021;Johnson et al 2023).…”

Section: Generating Knowledgeable Post Augmentations From Llmsmentioning

confidence: 99%

“…Despite the ineffectiveness of LLMs in this task, previous studies have demonstrated that LLMs exhibit strong language abilities, such as text comprehension, summarization, and sentiment analysis (Zhang et al 2023;Wang et al 2023b;Hsieh et al 2023), which can be used to distill useful knowledge for enhancing small models. On the one hand, we can leverage LLMs to generate post analyses (augmentations) from the aspects of semantic, sentiment, and linguistic, which are key factors for personality detection (Fang et al 2023). In this way, we can use the augmented post information to learn better post embeddings for personality detection.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

LLM vs Small Model? Large Language Model Based Text Augmentation Enhanced Personality Detection Model

Hu,

He,

Wang

et al. 2024

AAAI

View full text Add to dashboard Cite

Personality detection aims to detect one's personality traits underlying in social media posts. One challenge of this task is the scarcity of ground-truth personality traits which are collected from self-report questionnaires. Most existing methods learn post features directly by fine-tuning the pre-trained language models under the supervision of limited personality labels. This leads to inferior quality of post features and consequently affects the performance. In addition, they treat personality traits as one-hot classification labels, overlooking the semantic information within them. In this paper, we propose a large language model (LLM) based text augmentation enhanced personality detection model, which distills the LLM's knowledge to enhance the small model for personality detection, even when the LLM fails in this task. Specifically, we enable LLM to generate post analyses (augmentations) from the aspects of semantic, sentiment, and linguistic, which are critical for personality detection. By using contrastive learning to pull them together in the embedding space, the post encoder can better capture the psycho-linguistic information within the post representations, thus improving personality detection. Furthermore, we utilize the LLM to enrich the information of personality labels for enhancing the detection performance. Experimental results on the benchmark datasets demonstrate that our model outperforms the state-of-the-art methods on personality detection.

show abstract

Section: Generating Knowledgeable Post Augmentations From Llmsmentioning

confidence: 99%