Search papers, labs, and topics across Lattice.
This paper introduces FTA-Mem, a novel memory framework designed for low-density long-term dialogue in emotional-support agents, addressing the challenges of incomplete turns and evolving user states. By utilizing Boundary-preserving Window Segmentation (BWS) to create coherent situation fragments and Fact-Time-Affect Memory Units (FTA Units) that integrate factual, temporal, and affective information, FTA-Mem enhances the synthesis of structured context for response generation. Experimental results on the ES-MemEval and LoCoMo benchmarks demonstrate significant improvements in long-term memory question answering, achieving an F1 score of 0.3871 and a BERTScore of 0.6668, while effectively balancing evidence preservation and construction cost.
FTA-Mem achieves a remarkable balance between evidence preservation and construction cost, revolutionizing memory management in emotional-support dialogue systems.
Long-term emotional-support agents require memory mechanisms for personalized understanding across sessions. However, emotional-support dialogue is often low-density: turns are incomplete, evidence is scattered, and user states evolve over time. Existing memory methods usually rely on fixed units, such as turn-level notes or session summaries, which may lose details or introduce redundant noise. We propose FTA-Mem, a structured memory framework for low-density long-term dialogue. FTA-Mem uses Boundary-preserving Window Segmentation (BWS) to form coherent situation fragments, and constructs Fact-Time-Affect Memory Units (FTA Units) that jointly encode factual content, temporal grounding, and affective context. Retrieved units are then synthesized into structured context for answer generation. Experiments on ES-MemEval and LoCoMo show that FTA-Mem improves overall long-term memory question answering across benchmarks with different information-density characteristics. On ES-MemEval, FTA-Mem achieves 0.3871 F1 and 0.6668 BERTScore. Further analysis shows that situation-level FTA construction better balances evidence preservation and construction cost than coarse session-level or overly fine-grained turn-pair construction, providing an effective granularity trade-off for long-term dialogue memory.
