# PersonaDrift: A Benchmark for Temporal Anomaly Detection in Language-Based Dementia Monitoring

arXiv: https://arxiv.org/abs/2511.16445 · HTML: https://mlsanity.com/paper/2511.16445/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 17 of its 40 references that are in the map.
- First posted on arXiv 20 Nov 2025.

## Builds on (most cited first, 17 of 17)

- [DialogueRNN: An Attentive RNN for Emotion Detection in Conversations](https://mlsanity.com/paper/1811.00405/) (2018)
- [User Behavior Simulation with Large Language Model based Agents](https://mlsanity.com/paper/2306.02552/) (2023)
- [Simulating Classroom Education with LLM-Empowered Agents](https://mlsanity.com/paper/2406.19226/) (2024)
- [Roleplay-doh: Enabling Domain-Experts to Create LLM-simulated Patients via Eliciting and Adhering to Principles](https://mlsanity.com/paper/2407.00870/) (2024)
- [Large Language Models for Forecasting and Anomaly Detection: A Systematic Literature Review](https://mlsanity.com/paper/2402.10350/) (2024)
- [Less is More: Learning to Refine Dialogue History for Personalized Dialogue Generation](https://mlsanity.com/paper/2204.08128/) (2022)
- [Artificial Intelligence, speech and language processing approaches to monitoring Alzheimer's Disease: a systematic review](https://mlsanity.com/paper/2010.06047/) (2020)
- [How Far are LLMs from Being Our Digital Twins? A Benchmark for Persona-Based Behavior Chain Simulation](https://mlsanity.com/paper/2502.14642/) (2025)
- [Temporal Modeling Matters: A Novel Temporal Emotional Modeling Approach for Speech Emotion Recognition](https://mlsanity.com/paper/2211.08233/) (2022)
- [Adaptive Memory Networks with Self-supervised Learning for Unsupervised Anomaly Detection](https://mlsanity.com/paper/2201.00464/) (2022)
- [LLM Roleplay: Simulating Human-Chatbot Interaction](https://mlsanity.com/paper/2407.03974/) (2024)
- [Memory Augmented Generative Adversarial Networks for Anomaly Detection](https://mlsanity.com/paper/2002.02669/) (2020)
- [EmoUS: Simulating User Emotions in Task-Oriented Dialogues](https://mlsanity.com/paper/2306.01579/) (2023)
- [Coherent Dialogue with Attention-based Language Models](https://mlsanity.com/paper/1611.06997/) (2016)
- [A Self-Attentive Emotion Recognition Network](https://mlsanity.com/paper/1905.01972/) (2019)
- [A Systematic Review of NLP for Dementia -- Tasks, Datasets and Opportunities](https://mlsanity.com/paper/2409.19737/) (2024)
- [Adaptive Anomaly Detection in the Presence of Concept Drift: Extended Report](https://mlsanity.com/paper/2506.15831/) (2025)

## Abstract (arXiv)

People living with dementia (PLwD) often show gradual shifts in how they communicate, becoming less expressive, more repetitive, or drifting off-topic in subtle ways. While caregivers may notice these changes informally, most computational tools are not designed to track such behavioral drift over time. This paper introduces PersonaDrift, a synthetic benchmark designed to evaluate machine learning and statistical methods for detecting progressive changes in daily communication, focusing on user responses to a digital reminder system. PersonaDrift simulates 60-day interaction logs for synthetic users modeled after real PLwD, based on interviews with caregivers. These caregiver-informed personas vary in tone, modality, and communication habits, enabling realistic diversity in behavior. The benchmark focuses on two forms of longitudinal change that caregivers highlighted as particularly salient: flattened sentiment (reduced emotional tone and verbosity) and off-topic replies (semantic drift). These changes are injected progressively at different rates to emulate naturalistic cognitive trajectories, and the framework is designed to be extensible to additional behaviors in future use cases. To explore this novel application space, we evaluate several anomaly detection approaches, unsupervised statistical methods (CUSUM, EWMA, One-Class SVM), sequence models using contextual embeddings (GRU + BERT), and supervised classifiers in both generalized and personalized settings. Preliminary results show that flattened sentiment can often be detected with simple statistical models in users with low baseline variability, while detecting semantic drift requires temporal modeling and personalized baselines. Across both tasks, personalized classifiers consistently outperform generalized ones, highlighting the importance of individual behavioral context.