analiza aiqbench
This research introduces a schema-aware split learning framework that leverages an LLM to harmonize heterogeneous mental health survey data while preserving privacy. By partitioning the model between client devices and a central server, the approach significantly reduces local computational costs and outperforms standard federated learning methods on diverse datasets. The study demonstrates a viable path for collaborative AI training across institutions with varying data schemas without compromising sensitive information.
LLM-Based Schema-Aware Split Learning for Privacy-Preserving Mental Distress Prediction Across Heterogeneous Surveys
arXiv:2609.15871v1 Announce Type: cross Abstract: Rising societal and lifestyle complexity has been linked to a growing prevalence of mental distress worldwide. Educational institutions, workplaces, clinics, etc. collect large volumes of mental health survey data to understand and reduce this burden. Collaborative analysis of such data could yield effective generalizable predictive models. Privacy constraints and varied survey designs (i.e., different questions, scales, and formats) hinder direct integration. We propose a schema-aware split learning (SL) framework that preserves privacy, using a large language model (LLM) as a shared semantic encoder to harmonize heterogeneous survey schemas across institutions. We serialize each survey record into a natural-language description, unifying disparate survey schemas into a common format. The LLM is fine-tuned for mental distress assessment via Low-Rank Adaptation (LoRA) and partitioned across client and server. Clients retain the raw survey responses locally and run only a lightweight front-end, so original records never leave the institution that collected them. The resource-intensive backbone runs on the server, minimizing client-side computation. Using LLaMA-3.2-3B-Instruct, the framework attains an average ANLS of 0.708 with only 2,000 training samples, surpasses federated learning (FL) in eight of nine settings, and cuts per-client computation by three orders of magnitude, while generalizing to unseen datasets. Overall, it enables accurate, privacy-preserving, and resource-efficient collaborative learning from heterogeneous mental health survey data.
Abstract:Rising societal and lifestyle complexity has been linked to a growing prevalence of mental distress worldwide. Educational institutions, workplaces, clinics, etc. collect large volumes of mental health survey data to understand and reduce this burden. Collaborative analysis of such data could yield effective generalizable predictive models. Privacy constraints and varied survey designs (i.e., different questions, scales, and formats) hinder direct integration. We propose a schema-aware split learning (SL) framework that preserves privacy, using a large language model (LLM) as a shared semantic encoder to harmonize heterogeneous survey schemas across institutions. We serialize each survey record into a natural-language description, unifying disparate survey schemas into a common format. The LLM is fine-tuned for mental distress assessment via Low-Rank Adaptation (LoRA) and partitioned across client and server. Clients retain the raw survey responses locally and run only a lightweight front-end, so original records never leave the institution that collected them. The resource-intensive backbone runs on the server, minimizing client-side computation. Using LLaMA-3.2-3B-Instruct, the framework attains an average ANLS of 0.708 with only 2,000 training samples, surpasses federated learning (FL) in eight of nine settings, and cuts per-client computation by three orders of magnitude, while generalizing to unseen datasets. Overall, it enables accurate, privacy-preserving, and resource-efficient collaborative learning from heterogeneous mental health survey data.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
arXiv:2609.15871 [cs.LG]
(or arXiv:2609.15871v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.15871 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 312 cuvinte
Subiecte conexe
Claude Code’s Next Era — Thariq Shihipar, Anthropic
Latent Space · 29 sept. 2026
Links #3: 2026/06 Part 1
LessWrong · 15 iun. 2026
BettiSplit: Topology-Guided Privacy-Aware Split Learning Against Feature Inversion and Gradient Leakage
arXiv · 28 iul. 2026
[AINews] not much happened today
Latent Space · 3 oct. 2026