意识到上下文的层次关注抽象的对话总结.
Niya Yang1,2, Ye Wang3, Yichen Qi4
1Department of Information Security Technology, Jilin Police College, Changchun, 130123, China.
本研究介绍了一种层次上背景感知注意 (HCAtt) 网络,用于抽象对话总结. 在对话中,HCAtt有效地捕捉了多层次的背景,在基准数据集上表现优于现有的方法.
科学领域:
- 自然语言处理自然语言处理.
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 抽象的对话总结对于凝聚对话至关重要.
- 社会对话包含复杂的现象,如圆和话题转移.
- 现有的变压器模型很难有效地利用多层次的上下文信息.
研究的目的:
- 提出一个新的等级上下文意识注意力 (HCAtt) 网络.
- 通过结合细分层面和发言层面的上下文来增强对话总结.
- 改进对话数据中复杂的依赖关系的建模.
主要方法:
- 开发了一个层次上背景感知注意 (HCAtt) 网络.
- 在变压器框架内集成的细分级和发言级上下文信息.
- 在查询和关键转换过程中分层集成的上下文级别.
主要成果:
- 在基准数据集 (SAMSum,DialogSum,AMI) 上,HCAtt网络表现出卓越的性能.
- 该模型有效地捕捉了对话数据中的复杂依赖关系和上下文关系.
- 在抽象对话总结任务中表现优于现有方法.
结论:
- 对于抽象的对话总结,HCAtt网络是有效的.
- 纳入层次上下文可以显著改善对话的理解.
- 拟议的方法解决了传统变压器模型在处理对话复杂性的局限性.
更多相关视频
09:27Using Eye Movements Recorded in the Visual World Paradigm to Explore the Online Processing of Spoken Language
Published on: October 13, 2018
08:45Mapping Cortical Dynamics Using Simultaneous MEG/EEG and Anatomically-constrained Minimum-norm Estimates: an Auditory Attention Example
Published on: October 24, 2012
相关概念视频
Techniques of Therapeutic Communication II: Focusing, Paraphrasing, and Summarizing
This therapeutic technique can also be used when a patient brings up pertinent information during a health-related conversation. The...
Subconsciousness and No Awareness
An illustrative example of subconscious processing is its role in problem-solving. Often, individuals...
The Anchoring-and-Adjustment Heuristic
High-Level and Low-Level Awareness
Cerebral Hemispheres
Higher Mental Functions of the Brain: Language
Language formation and comprehension take place in the dominant hemisphere. The dominant hemisphere is responsible for understanding the meaning of spoken, written, or sign language, as well as the ability to communicate. For most people, the left hemisphere is the dominant one. The right hemisphere, then, gives tone and emotional context to the...
