分离检索与生成，让音乐推荐更有依据。
Two Views, One Voice: Evidence-Grounded Conversational Music Recommendation
- 分两阶段处理：先精准检索，再基于证据生成回复
- 在盲评中解释质量排名第二，接近最佳表现
- 适合需要可解释推荐系统的研发团队
传统对话推荐系统将检索与生成混在同一文本界面中，导致实体线索随对话意图演变而模糊，影响解释可信度。我们在 ACM RecSys Challenge 2026 的 Blind-B 行业赛道中提出第三名解决方案。团队 'swyoo' 将检索与响应解耦为独立管道，通过排序后的曲目和元数据严格连接。检索采用混合词法-稠密池，结合微调的 Qwen 8B 适配器驱动的任务适配池。候选曲目经 LightGBM 校准后，送入证据引导的提出-分配-选择（PAS）框架以结构化生成。该系统在最终盲评中解释质量榜上位列第二。结果表明：(i) 分离检索与生成能同时保持曲库线索与对话意图流畅性；(ii) 通过显式证据分配结构化生成，是实现近最佳解释可靠性的关键。
原文摘要 · Abstract (English)
Traditional conversational recommenders entangle retrieval and response generation within a single text interface, so exact entity cues fade as the dialogue's intent evolves, which compromises explanation credibility. We address this within the ACM RecSys Challenge 2026, which mandates both top-20 ranking and evidence-grounded response generation. This paper presents the third-place solution by team "swyoo" for the Blind-B industry track. We decouple retrieval and response into separate pipelines connected strictly via ranked tracks and metadata. Retrieval combines a hybrid lexical-dense pool for exact matching with a task-adapted pool driven by fine-tuned Qwen 8B adapters. Candidates are calibrated via LightGBM, then routed to an evidence-grounded propose-assign-select (PAS) framework to structure responses. This system also ranked second on the explanation-quality leaderboard in the final blind evaluation. Our findings demonstrate that: (i) isolating retrieval and response preserves both catalog cues and fluid intent; (ii) structuring generation via explicit evidence assignment is key to this near-best-in-class explanation reliability.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。