English(EN) DeepOpen: Decision-Oriented Intent Classification on BANKING77 and CLINC150

DeepOpen框架在银行和意图分类任务上实现了高精度

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-26 04:00

研究人员推出DeepOpen，一个为BANKING77和CLINC150等特定数据集量身定制的面向决策的意图分类框架。该系统使用单一的带锐度感知最小化（SAM）的英文分类器在BANKING77上达到了94.12%的高准确率。对于CLINC150，Laya专家和DeBERTa-v3-large分类器的概率融合达到了98.02%的准确率。该论文详细介绍了配置，强调了性能指标，并讨论了潜在的选择效应和历史执行路径差异，而不是声称新的算法优势。 AI

影响这项研究展示了在特定数据集上提高意图分类准确性，可能有利于需要细致用户理解的应用。

排序理由该集群包含一篇学术论文，详细介绍了新框架及其在特定数据集上的性能。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 13 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇学术论文，详细介绍了新框架及其在特定数据集上的性能。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Shuang Song, Muhammad Luqman Naseem, Zhongqiao Yuan · 2026-10-26 04:00

DeepOpen：面向决策的意图分类在BANKING77和CLINC150上的应用

arXiv:2610.31104v2 Announce Type: replace Abstract: We present DeepOpen, a decision-Oriented intent-classification framework with dataset- specific systems for BANKING77 and CLINC150. On BANKING77, a single English classi- fier continued with sharpness-aware minimization (SAM) ac…

本文如何被排名

报道来源 [1]

DeepOpen：面向决策的意图分类在BANKING77和CLINC150上的应用

相关实体

相关话题