A Proactive Multi-Agent Dialogue Framework for Assessing Social Language Disorder Traits in Autism
A new framework called TPA has been developed to enhance the assessment of Social Language Disorder traits in individuals with autism. This proactive multi-agent dialogue system improves the efficiency of language assessments by strategically selecting questioning methods. The framework has shown significant improvements in trait coverage and diagnostic efficiency compared to traditional methods.
- ▪The TPA framework outperforms traditional dialogue methods, achieving 82.1% SLD trait coverage.
- ▪It demonstrates a 16.6% improvement over automated replay of real clinical dialogues conducted by trained clinicians.
- ▪The system was validated across three independent experiments using data from 35 patients.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Computation and Language arXiv:2605.22993 (cs) [Submitted on 21 May 2026] Title:A Proactive Multi-Agent Dialogue Framework for Assessing Social Language Disorder Traits in Autism Authors:Chuanbo Hu, Minglei Yin, Bin Liu, Wenqi Li, Lynn K. Paul, Shuo Wang, Xin Li View a PDF of the paper titled A Proactive Multi-Agent Dialogue Framework for Assessing Social Language Disorder Traits in Autism, by Chuanbo Hu and 6 other authors View PDF HTML (experimental) Abstract:Characteristic linguistic behaviors associated with Social Language Disorder (SLD) in autism spectrum disorder, including echoic repetition, pronoun displacement, and stereotyped media quoting, are largely absent from spontaneous conversation and only emerge under specific conversational conditions.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
