ES-MemEval: Benchmarking Conversational Agents on Personalized Long-Term Emotional Support

TL;DR

This paper introduces ES-MemEval, a benchmark for evaluating long-term memory capabilities of conversational agents in emotional support, along with EvoEmo, a dataset capturing implicit, evolving user disclosures, revealing current models' strengths and limitations.

Contribution

It presents a new benchmark and dataset specifically designed for assessing and advancing long-term memory in emotional support dialogue systems.

Findings

Explicit memory reduces hallucinations and improves personalization.

Retrieval-augmented models enhance factual accuracy but struggle with temporal dynamics.

Current models show limitations in handling evolving user states.

Abstract

Large Language Models (LLMs) have shown strong potential as conversational agents. Yet, their effectiveness remains limited by deficiencies in robust long-term memory, particularly in complex, long-term web-based services such as online emotional support. However, existing long-term dialogue benchmarks primarily focus on static and explicit fact retrieval, failing to evaluate agents in critical scenarios where user information is dispersed, implicit, and continuously evolving. To address this gap, we introduce ES-MemEval, a comprehensive benchmark that systematically evaluates five core memory capabilities: information extraction, temporal reasoning, conflict detection, abstention, and user modeling, in long-term emotional support settings, covering question answering, summarization, and dialogue generation tasks. To support the benchmark, we also propose EvoEmo, a multi-session dataset for personalized long-term emotional support that captures fragmented, implicit user disclosures and evolving user states. Extensive experiments on open-source long-context, commercial, and retrieval-augmented (RAG) LLMs show that explicit long-term memory is essential for reducing hallucinations and enabling effective personalization. At the same time, RAG improves factual consistency but struggles with temporal dynamics and evolving user states. These findings highlight both the potential and limitations of current paradigms and motivate more robust integration of memory and retrieval for long-term personalized dialogue systems.