EUVIMEDEUVIMED entdecken
Crossref

Detecting Emotional Dynamic Trajectories: An Evaluation Framework for Emotional Support in Language Models

Zhouxing Tan, Ruochong Xiong, Yulong Wan, Jinlong Ma, Hanlin Xue, Qichun Deng, Haifeng Jing, Zhengtong Zhang, Depei Liu, Shiyuan Luo, Junfei Liu

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (3) · 2074-2082

DOI: 10.1609/aaai.v40i3.37189 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Zhouxing Tan und 10 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Emotional support is a core capability in human-AI interaction, with applications including psychological counseling, role play, and companionship. However, existing evaluations of large language models (LLMs) often rely on short, static dialogues and fail to capture the dynamic and long-term nature of emotional support. To overcome this limitation, we shift from snapshot-based evaluation to trajectory-based assessment, adopting a user-centered perspective that evaluates models based on their ability to improve and stabilize user emotional states over time. Our framework constructs a large-scale benchmark consisting of 328 emotional contexts and 1,152 disturbance events, simulating realistic emotional shifts under evolving dialogue scenarios. To encourage psychologically grounded responses, we constrain model outputs using validated emotion regulation strategies such as situation selection and cognitive reappraisal. User emotional trajectories are modeled as a first-order Markov process, and we apply causally-adjusted emotion estimation to obtain unbiased emotional state tracking. Based on this framework, we introduce three trajectory-level metrics: Baseline Emotional Level (BEL), Emotional Trajectory Volatility (ETV), and Emotional Centroid Position (ECP). These metrics collectively capture user emotional dynamics over time and support comprehensive evaluation of long-term emotional support performance of LLMs. Extensive evaluations across a diverse set of LLMs reveal significant disparities in emotional support capabilities and provide actionable insights for model development.

Zitieren

APA

Tan, Z., Xiong, R., Wan, Y., Ma, J., Xue, H., Deng, Q., Jing, H., Zhang, Z., Liu, D., Luo, S., & Liu, J. (2026). Detecting Emotional Dynamic Trajectories: An Evaluation Framework for Emotional Support in Language Models. Proceedings of the AAAI Conference on Artificial Intelligence, 40(3), 2074-2082. https://doi.org/10.1609/aaai.v40i3.37189

Vancouver

Tan Z, Xiong R, Wan Y, Ma J, Xue H, Deng Q, et al. Detecting Emotional Dynamic Trajectories: An Evaluation Framework for Emotional Support in Language Models. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(3):2074-2082. doi:10.1609/aaai.v40i3.37189

BibTeX
@article{Tan2026,
  author = {Tan, Zhouxing and Xiong, Ruochong and Wan, Yulong and Ma, Jinlong and Xue, Hanlin and Deng, Qichun and Jing, Haifeng and Zhang, Zhengtong and Liu, Depei and Luo, Shiyuan and Liu, Junfei},
  title = {Detecting Emotional Dynamic Trajectories: An Evaluation Framework for Emotional Support in Language Models},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {3},
  pages = {2074--2082},
  doi = {10.1609/aaai.v40i3.37189}
}
Kostenlos starten →