English(EN) From Clicks to Intent: Cross-Platform Session Embeddings with LLM-Distilled Taxonomy for Financial Services Recommendations

LLM提炼的分类法改进金融服务推荐

作者 PulseAugur 编辑部 · [2 个来源] · 2026-06-24 18:18

研究人员开发了一个新框架，通过弥合登录前网页互动与已认证应用内体验之间的差距，来改善金融服务的个性化。该系统使用自监督Transformer从点击流创建会话嵌入，并使用基于LLM的管道生成可解释的意图标签。这种双重方法增强了主页图块排名和用户转化预测等量化任务，同时还以低延迟提供了定性理解。 AI

影响这项研究通过更好地理解跨不同平台的用��意图，可能带来更有效和个性化的金融服务推荐。

排序理由该集群包含一篇学术论文，详细介绍了AI应用的新方法。

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.AI TIER_1 English(EN) · Dianjing Fan, Yao Li, Kyaw Hpone Myint, Dwipam Katariya, Alexandre G. R. Day, Pranab Mohanty, Giri Iyengar · 2026-06-26 04:00

From Clicks to Intent: Cross-Platform Session Embeddings with LLM-Distilled Taxonomy for Financial Services Recommendations

arXiv:2606.26277v1 Announce Type: cross Abstract: Sequential user behavior modeling is widely adopted in industrial recommender systems; however, significant gaps remain in financial services, where pre-login web interactions and authenticated in-app experiences differ drasticall…
arXiv cs.IR (Information Retrieval) TIER_1 English(EN) · Giri Iyengar · 2026-06-24 18:18

From Clicks to Intent: Cross-Platform Session Embeddings with LLM-Distilled Taxonomy for Financial Services Recommendations

Sequential user behavior modeling is widely adopted in industrial recommender systems; however, significant gaps remain in financial services, where pre-login web interactions and authenticated in-app experiences differ drastically. Specifically, pre-login web users typically exp…