dorsal/arxiv
View SchemaES-Mem: Event Segmentation-Based Memory for Long-Term Dialogue Agents
| Authors | Huhai Zou, Tianhao Sun, Chuanjiang He, Yu Tian, Zhenyang Li, Li Jin, Nayu Liu, Jiang Zhong, Kaiwen Wei |
|---|---|
| Categories | |
| ArXiv ID | 2601.07582vv2 |
| URL | https://arxiv.org/abs/2601.07582 |
| License | http://creativecommons.org/licenses/by/4.0/ |
Abstract
Memory is critical for dialogue agents to maintain coherence and enable continuous adaptation in long-term interactions. While existing memory mechanisms offer basic storage and retrieval capabilities, they are hindered by two primary limitations: (1) rigid memory granularity often disrupts semantic integrity, resulting in fragmented and incoherent memory units; (2) prevalent flat retrieval paradigms rely solely on surface-level semantic similarity, neglecting the structural cues of discourse required to navigate and locate specific episodic contexts. To mitigate these limitations, drawing inspiration from Event Segmentation Theory, we propose ES-Mem, a framework incorporating two core components: (1) a dynamic event segmentation module that partitions long-term interactions into semantically coherent events with distinct boundaries; (2) a hierarchical memory architecture that constructs multi-layered memories and leverages boundary semantics to anchor specific episodic memory for precise context localization. Evaluations on two memory benchmarks demonstrate that ES-Mem yields consistent performance gains over baseline methods. Furthermore, the proposed event segmentation module exhibits robust applicability on dialogue segmentation datasets.
{
"annotation_id": "868d494e-cb23-4f8c-a6b3-8265c2848f26",
"date_created": "2026-02-17T05:53:12.641000Z",
"date_modified": "2026-02-17T05:53:12.641000Z",
"file_hash": "d6769b5531569d10fada3de9ad37efff3b5659710c4b9caf45c9708980f878c4",
"private": false,
"record": {
"abstract": "Memory is critical for dialogue agents to maintain coherence and enable continuous adaptation in long-term interactions. While existing memory mechanisms offer basic storage and retrieval capabilities, they are hindered by two primary limitations: (1) rigid memory granularity often disrupts semantic integrity, resulting in fragmented and incoherent memory units; (2) prevalent flat retrieval paradigms rely solely on surface-level semantic similarity, neglecting the structural cues of discourse required to navigate and locate specific episodic contexts. To mitigate these limitations, drawing inspiration from Event Segmentation Theory, we propose ES-Mem, a framework incorporating two core components: (1) a dynamic event segmentation module that partitions long-term interactions into semantically coherent events with distinct boundaries; (2) a hierarchical memory architecture that constructs multi-layered memories and leverages boundary semantics to anchor specific episodic memory for precise context localization. Evaluations on two memory benchmarks demonstrate that ES-Mem yields consistent performance gains over baseline methods. Furthermore, the proposed event segmentation module exhibits robust applicability on dialogue segmentation datasets.",
"arxiv_id": "2601.07582",
"authors": [
"Huhai Zou",
"Tianhao Sun",
"Chuanjiang He",
"Yu Tian",
"Zhenyang Li",
"Li Jin",
"Nayu Liu",
"Jiang Zhong",
"Kaiwen Wei"
],
"categories": [
"cs.CL",
"cs.AI"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"title": "ES-Mem: Event Segmentation-Based Memory for Long-Term Dialogue Agents",
"url": "https://arxiv.org/abs/2601.07582",
"version": "v2"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "4d4cd847-17a4-492f-94bf-1d8ec1496d6d",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}