Since its inception, this workshop series has served as a multidisciplinary forum for researchers to explore the intersection of cognitive science, linguistics, and artificial intelligence. Our goal is to foster the development of dialogue agents that do not merely mimic human speech patterns but understand and replicate the underlying cognitive processes that drive human interaction. The 10th edition of LaCATODA was accepted to AAAI 2026, one of the top AI conferences, and was held as a half-day workshop at Singapore EXPO. As Large Language Models (LLMs) continue to dominate the field of Natural Language Processing, the 2026 edition of LACATODA highlights a critical shift. This year's contributions focus on moving beyond surface-level fluency to address complex challenges in personality consistency, emotional intelligence, metaphorical reasoning, and social mediation. The workshop received 21 submissions from researchers worldwide. Following a rigorous peer-review process, one paper was withdrawn, and 14 papers were accepted for presentation (7 oral presentations and 7 posters), resulting in an acceptance rate of 66%. Two papers were subsequently excluded from the proceedings due to the authors' absence. The technical program reflects the diverse landscape of modern dialogue research. Key themes include: • Personality and Emotion: Research explored how LLMs manifest Big Five traits across contexts [1] and their alignment with intended personality prompts [6]. Innovative work also examined emotion-conditioned generation using visual novel data [10]. • Dialogue Structure and Intent: New approaches were proposed for rethinking dialogue disentanglement through dialogue-level assignment [2] and utilizing ontology-guided intent routing for multi-agent systems [11]. • Reasoning and Refinement: Methodologies were introduced to improve multi-step inference, such as using uncertainty-guided latent space search [7] and multi-agent frameworks for response refinement [3]. • Social Nuance and Figurative Language: Contributors presented a novel neuro-symbolic approach to irony detection based on structural components [12]. Further social studies investigated LLMs as dispute mediators [8].
• Security and Utility: Research addressed Japanese phishing detection [
We were honored to host an invited talk by Dr. Rui Mao (Nanyang Technological University) titled “Metaphorical Cognition and Its Computational Practices.” Dr. Mao presented the MetaPro framework, demonstrating how metaphorical cognition serves as a core mechanism for organizing human concepts. His talk highlighted how large-scale metaphor processing can uncover latent cognitive patterns in discourse across domains such as finance and mental health.
We would like to express our deepest gratitude to the Program Committee for their invaluable time and detailed feedback. We also thank the authors for their high-quality contributions and the participants for their engaging discussions that continue to drive this community forward.
Program Committee • Don Divin Anameta, Hokkaido University, Japan • Nor Saiful Azam Bin Nor Azmi, Kitami Institute of Technology, Japan • Lin Chia, Kanazawa Gakuin University, Japan • Abu Nowhash Chowdhury, International Islamic University Chittagong, Bangladesh • Siaw-Fong Chung, National Chengchi University, Taiwan • Karolina Drozdz, IDEAS RI, Poland • Kacper Dudzic, IDEAS RI/ Adam Mickiewicz University, Poland • Juuso Eronen, Prefectural University of Kumamoto, Japan • Magdalena Igras-Cybulska, AGH, Poland • Khalifah Mujahid Jamal A, Saudi Arabia • Pawel Lempa, Cracow University of Technology, Poland • Chia Zheng Lin, Kanazawa Gakuin University, Japan • Da Li, Kyoto Sangyo University, Japan • Tanjim Mahmud, Rangamati Science and Technology University, Bangladesh • Michal Mazur, Ritsumeikan Asia Pacific University, Japan • Karol Nowakowski, University of Yamanashi, Japan • Michal Ptaszynski, Kitami Institute of Technology, Japan • Rafal Rzepka, Hokkaido University, Japan • Hiroki Sakaji, Hokkaido University, Japan • Ryoma Shinto, Hokkaido University, Japan • Daiki Shirafuji, Mitsubishi Electric, Japan • Marcin Skowron, Deeprsearch GmbH & Austrian Research Institute for AI, Austria • Joanna Staszkow, Orange, Poland • Sho Takishita, Fujitsu, Japan • Lu Wang, Kitami Institute of Technology, Japan • Zanwei Wang, Nihon University, Japan • Liu Zhenzhen, Kitami Institute of Technology, Japan [11] N. Wiratunga, V. A. Wijayasekara, I. Nkisi-Orji, P. Salimi, K. Martin and C. Bolaños. CARE: Ontology-Guided Intent Routing for Multi-Agent LLM-Based Dialogue Systems. In the Proceedings of The 10th Linguistic and Cognitive Approaches To Dialog Agents (LaCATODA 2026) Workshop at AAAI 2026 Conference, pp. 170–181, Singapore, 2026. [12] H. Shigenobu, M. Ptaszynski, S. Dan, F. Masui, Y. Uchida, and R. Rzepka. A Novel Neuro-symbolic Approach to Irony Detection Based on Structural Components of Ironic Statements. In the Proceedings of The 10th Linguistic and Cognitive Approaches To Dialog Agents (LaCATODA 2026) Workshop at AAAI 2026 Conference, pp. 181–193, Singapore, 2026.