Personality-Driven Social Multimedia Content Recommendation

Yang, Qi; Nikolenko, Sergey I.; Huang, Alfred; Farseev, Aleksandr

doi:10.1145/3503161.3548769

Cited by 10 publications

(5 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Techniques such as feature importance analysis and decision tree approximations were shown to provide valuable insights into model reasoning, albeit with limitations in capturing the full complexity of LLMs [38,26,39,40]. Another approach that garnered attention involves the integration of interpretability directly into the model architecture, through methods like attention mechanisms and transparent layers designed to offer more direct insight into the decision-making process [38,41,34,36,32]. Despite these advancements, challenges remain in achieving a satisfactory level of interpretability without compromising model performance, underscoring the ongoing need for innovative solutions.…”

Section: Methodologies For Enhancing Model Interpretabilitymentioning

confidence: 99%

“…The development and implementation of frameworks for model explainability have been a focal point of research efforts. Such frameworks aim to provide users with understandable explanations for the decisions made by LLMs, often through the generation of natural language explanations or visualizations of the decision path [38,36,28,41,63,64,65]. It was observed that those frameworks significantly improve user trust and satisfaction, particularly in domains where understanding AI decisions is critical [65,27,66,56,57].…”

Section: Framework For Model Explainabilitymentioning

confidence: 99%

“…It was observed that those frameworks significantly improve user trust and satisfaction, particularly in domains where understanding AI decisions is critical [65,27,66,56,57]. However, the effectiveness of different explainability frameworks varies, with some studies noting that the quality and utility of explanations depend heavily on the context in which they are provided, including those in LLM explainability benchmarks [38,41,67,65]. Additionally, there has been an emphasis on the adaptability of explainability frameworks, allowing them to be tailored to specific user needs and levels of expertise [64,6,68,59,69,70,71,72,73].…”

Section: Framework For Model Explainabilitymentioning

confidence: 99%

See 2 more Smart Citations

Measuring the Interpretability and Explainability of Model Decisions of Five Large Language Models

Fujiwara,

Sasaki,

Nakamura

et al. 2024

Preprint

View full text Add to dashboard Cite

This study conducts a comprehensive analysis of the interpretability and explainability of five leading Large Language Models (LLMs): TripoSR by Stability AI, Gemma-7b by Google, Mistral 7B by Mistral AI, Llama-2-7b by Meta, and GemMoE-Beta-1 by CrystalCare AI. Through a methodical evaluation encompassing both qualitative and quantitative benchmarks, we assess these models' capacity to make their decision-making processes understandable to humans. Our findings reveal significant variability in the models' ability to provide transparent reasoning and accurate, contextually relevant explanations across different contexts. Notably, TripoSR and GemMoE-Beta-1 demonstrated superior transparency, while Gemma-7b and Llama-2-7b excelled in the accuracy of their explanations. However, challenges in maintaining consistent interpretability and explainability across varying inputs and the need for enhanced adaptability to feedback highlight areas for future improvement. This research underscores the importance of interpretability and explainability in fostering trust and reliability in LLM applications, advocating for continued advancement in these areas to achieve more transparent, accountable, and user-centric AI systems. Directions for future research include the development of standardized evaluation methodologies and interdisciplinary approaches to enhance model transparency and user understanding.

show abstract

Section: Methodologies For Enhancing Model Interpretabilitymentioning

confidence: 99%

Section: Framework For Model Explainabilitymentioning

confidence: 99%

Section: Framework For Model Explainabilitymentioning

confidence: 99%

See 1 more Smart Citation

Measuring the Interpretability and Explainability of Model Decisions of Five Large Language Models

Fujiwara,

Sasaki,

Nakamura

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Personality detection aims to identify the personality traits implied in social media posts that offering a deeper insight into human behavior (Nutescu and Mocanu 2023), emotional processes (Lian, Liu, and Tao 2022), and mental health (Zanwar et al 2023). Besides, it can provide timely and objective support for downstream applications, such as human-computer interaction systems (Chien, Chen, and Chan 2022), virtual dialogue systems (Yang, Chen, and Narasimhan 2021), and recommendation systems (Yang et al 2022;Shen et al 2020).…”

Section: Introductionmentioning

confidence: 99%

Data Augmented Graph Neural Networks for Personality Detection

Zhu,

Xia,

et al. 2024

AAAI

View full text Add to dashboard Cite

Personality detection is a fundamental task for user psychology research. One of the biggest challenges in personality detection lies in the quantitative limitation of labeled data collected by completing the personality questionnaire, which is very time-consuming and labor-intensive. Most of the existing works are mainly devoted to learning the rich representations of posts based on labeled data. However, they still suffer from the inherent weakness of the amount limitation of labels, which potentially restricts the capability of the model to deal with unseen data. In this paper, we construct a heterogeneous personality graph for each labeled and unlabeled user and develop a novel psycholinguistic augmented graph neural network to detect personality in a semi-supervised manner, namely Semi-PerGCN. Specifically, our model first explores a supervised Personality Graph Neural Network (PGNN) to refine labeled user representation on the heterogeneous graph. For the remaining massive unlabeled users, we utilize the empirical psychological knowledge of the Linguistic Inquiry and Word Count (LIWC) lexicon for multi-view graph augmentation and perform unsupervised graph consistent constraints on the parameters shared PGNN. During the learning process of finite labeled users, noise-invariant learning on a large scale of unlabeled users is combined to enhance the generalization ability. Extensive experiments on three real-world datasets, Youtube, PAN2015, and MyPersonality demonstrate the effectiveness of our Semi-PerGCN in personality detection, especially in scenarios with limited labeled users.

show abstract

“…The impact of global marketing on revenue (Gutiérrez-Aragón et al, 2023; Taron et al, 2023; Yang et al, 2022) kele honey entrepreneur in Balangan Sembung Mengwi Bali can vary depending on a number of factors. Global marketing can help entrepreneurs (Chen et al, 2023) Kele honey to expand market access outside Bali.…”

Section: Introductionmentioning

confidence: 99%

Global Marketing to Increase Income of Kele Honey Entrepreneurs in Balangan Sembung Mengwi Bali

Laksmi,

Selamet,

Pasek Mangku

et al. 2023

IJSMR

View full text Add to dashboard Cite

This study aims to analyze the effect of global marketing on increasing the income of kele honey entrepreneurs in Balangan Sembung Village, Mengwi, Bali. A total of 42 respondents participated in the study. Global marketing is considered to have a major impact on revenue, so this study is important to examine the effect of global marketing on revenue. The hypothesis in this study states that global marketing has a positive and significant influence on the income of honey kele entrepreneurs in Balangan Sembung Village, Mengwi, Bali. The analysis was conducted to assess the extent to which global marketing affects the increase in income of honey kele entrepreneurs. The results of this study are expected to provide insight into the importance of global marketing in increasing the income of honey kele entrepreneurs in the region. The implications of these findings are expected to support the development of more effective marketing strategies to market kele honey products globally, having a positive impact on their future income.

show abstract

Personality-Driven Social Multimedia Content Recommendation

Cited by 10 publications

References 53 publications

Measuring the Interpretability and Explainability of Model Decisions of Five Large Language Models

Measuring the Interpretability and Explainability of Model Decisions of Five Large Language Models

Data Augmented Graph Neural Networks for Personality Detection

Global Marketing to Increase Income of Kele Honey Entrepreneurs in Balangan Sembung Mengwi Bali

Contact Info

Product

Resources

About