While Large Language Models (LLMs) have advanced open-domain dialogue systems, maintaining long-term consistency remains a challenge due to inherent limitations in long-context reasoning and the inefficiency of processing extensive raw text. Existing approaches typically rely on either unstructured memory storage, which is prone to information loss, or computationally expensive LLMs that incur high latency. To address these limitations, we propose G-Long, a graph-enhanced framework that utilizes a fine-tuned small Language Model (sLM) for structured triplet extraction and associative retrieval, significantly reducing operational costs. Furthermore, we introduce the novel attention-aware importance scoring mechanism that leverages the intrinsic cross-attention signals of a T5 summarizer to identify salient memories. Extensive experiments across diverse benchmarks demonstrate that G-Long achieves state-of-the-art performance in both response generation and memory retrieval, yielding performance gains of up to 9.8% in response quality on MSC and 40.8% in retrieval recall on LME, while significantly minimizing computational overhead.

翻译：尽管大规模语言模型（LLMs）推动了开放域对话系统的发展，但由于长上下文推理的内在局限性以及处理海量原始文本的低效率，维持长期一致性仍面临挑战。现有方法通常依赖非结构化记忆存储（易致信息丢失）或计算成本高昂的LLMs（导致高延迟）。为克服这些局限，我们提出G-Long——一个采用微调小型语言模型（sLM）进行结构化三元组提取与关联检索的图增强框架，显著降低了运算成本。此外，我们引入创新的注意力感知重要性评分机制，利用T5摘要器固有的交叉注意力信号识别关键记忆。跨多基准的大量实验表明，G-Long在响应生成与记忆检索两方面均达到顶尖性能：在MSC上响应质量提升高达9.8%，在LME上检索召回率提升40.8%，同时大幅降低计算开销。