CALLM: Understanding Cancer Survivors' Emotions and Intervention Opportunities via Mobile Diaries and Context-Aware Language Models

TL;DR

This work tackles how to monitor cancer survivors' day-to-day emotions with minimal user burden and deliver timely, personalized support. It introduces CALLM, a context-aware language-model framework that uses retrieval-augmented generation and personal diary history to infer emotional states and intervention opportunities from extremely brief mobile diary entries. Through a five-week EMA study with $N=407$ survivors and $24{,}183$ diary entries, CALLM uncovers robust context–emotion relationships (e.g., leisure contexts boost positive affect; administrative/cancer-related contexts elevate negative affect) and achieves strong predictive performance ($72.96\%$ for positive affect, $73.29\%$ for negative affect, $73.72\%$ for emotion regulation desire, $60.09\%$ for intervention availability), outperforming baselines. Post-hoc analyses show model confidence, diary length, and personalization materially influence accuracy, supporting deployment strategies that balance reliability and user burden. The findings offer a practical pathway to privacy-preserving, context-aware JITAI systems that leverage mobile diaries to tailor emotional support for cancer survivors.

Abstract

Cancer survivors face unique emotional challenges that impact their quality of life. Mobile diary entries provide a promising method for tracking emotional states, improving self-awareness, and promoting well-being outcome. This paper aims to, through mobile diaries, understand cancer survivors' emotional states and key variables related to just-in-time intervention opportunities, including the desire to regulate emotions and the availability to engage in interventions. Although emotion analysis tools show potential for recognizing emotions from text, current methods lack the contextual understanding necessary to interpret brief mobile diary narratives. Our analysis of diary entries from cancer survivors (N=407) reveals systematic relationships between described contexts and emotional states, with administrative and health-related contexts associated with negative affect and regulation needs, while leisure activities promote positive emotions. We propose CALLM, a Context-Aware framework leveraging Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) to analyze these brief entries by integrating retrieved peer experiences and personal diary history. CALLM demonstrates strong performance with balanced accuracies reaching 72.96% for positive affect, 73.29% for negative affect, 73.72% for emotion regulation desire, and 60.09% for intervention availability, outperforming language model baselines. Post-hoc analysis reveals that model confidence strongly predicts accuracy, with longer diary entries generally enhancing performance, and brief personalization periods yielding meaningful improvements. Our findings demonstrate how contextual information in mobile diaries can be effectively leveraged to understand emotional experiences, predict key states, and identify optimal intervention moments for personalized just-in-time support.

Figures (12)

• Figure 1: Distrubutions of word count and sentiment polarity across collected mobile diaries from N=407 cancer survivors.
• Figure 2: EMA interface for collecting emotional measures, desire to regulate emotions, and availability to do interventions.
• Figure 3: Emotional and Behavioral Measures Across Different Emotion Driver Categories.
• Figure 4: Emotional States and Intervention Availability Across Different Sentiment Categories
• Figure 5: Word clouds with high (above individual mean) affective scores reported.