数据质量很重要:使用RoBERTa-CNN检测社交媒体帖子上的自杀意图
概括
这项研究引入了一个新的深度学习模型,RoBERTa-CNN,用于检测在线帖子中的自杀意图. 该模型实现了98%的准确性,强调了数据质量在预防自杀研究中的重要性.
科学领域:
- 计算语言学计算语言学
- 人工智能的人工智能是人工智能.
- 心理健康技术 心理健康技术
背景情况:
- 自杀是全球主要的健康问题,需要先进的检测方法.
- 像Reddit这样的在线平台包含有价值的数据,用于识别有风险的个人.
- 现有的方法可能无法完全捕捉文本中表达的自杀意念的细微差别.
研究的目的:
- 提出和评估一种新的深度学习模型,用于识别Reddit帖子中的自杀意图.
- 为了利用 RoBERTa 和 CNN 的优势进行增强的文本分析.
- 调查数据质量对自杀检测模型性能的影响.
主要方法:
- 利用了罗伯塔-CNN深度学习模型,将罗伯塔用于语义理解和CNN用于模式识别.
- 在自杀和抑郁症检测数据集上训练和评估模型.
- 实施数据清理技术,包括手动清理和OpenAI API,以提高文本数据质量.
主要成果:
- 罗伯塔-CNN模型的平均精度为98%,标准偏差为0.0009.
- 证明数据质量显著影响检测模型的稳定性和性能.
- 数据预处理步骤对于优化模型训练至关重要.
结论:
- 罗伯塔-CNN模型显示,在线文本中早期发现自杀意图具有显著的前景.
- 高质量,清洁的数据对于开发有效的AI驱动的心理健康工具至关重要.
- 这种方法提供了一个可扩展的解决方案,用于在线心理健康环境中的监测和干预.
更多相关视频
08:53Integrating Computerized Linguistic and Social Network Analyses to Capture Addiction Recovery Capital in an Online Community
Published on: May 31, 2019
5.9K
07:31Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
8.3K
相关概念视频
Self-Presentation: Self-Monitoring and Self-Handicapping
45.3K
People can go to great lengths to protect their self-image and present themselves in ways that they want others to see them. Sociologist Erving Goffman presented the idea that a person is like an actor on a stage. Calling his theory dramaturgy, Goffman believed that we use “impression management” to present ourselves to others as we hope to be perceived. Each situation is a new scene, and individuals perform different roles depending on who is present (Goffman, 1959). Think about...
45.3K
Social Foundations of Self IV: Self in Digital Communication
273
Since the early 2000s, computer-mediated communication (CMC) has grown rapidly, playing a crucial role in self-development. A key distinction between CMC and real-life interactions is the lack of a physically present partner. This absence makes non-verbal cues such as facial expressions, body language, and paralinguistic signals unavailable in CMC platforms like email, instant messaging, or social media. The lack of these cues can create ambiguity and complicate how feedback is interpreted.The...
273
Strategies of Self-Presentation III: Self-Monitoring
416
Self-monitoring is a central construct in understanding individual differences in self-presentation strategies across social contexts. It refers to how individuals observe, regulate, and control their expressive behavior and self-presentation following situational cues. Self-monitoring reflects a person's sensitivity to social appropriateness and willingness to adapt behavior to fit varying interpersonal demands.High vs. Low Self-Monitoring IndividualsIndividuals high in self-monitoring are...
416
