arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.23376cs.HC
收藏

通过持续反馈循环实现动态个性化:交互式AI系统中

Dynamic Personalization Through Continuous Feedback Loops in Interactive AI Systems

bitc.edu.cn(北京信息科技大学)

Liu He

AI总结:

本研究提出通过持续反馈循环实现动态个性化,提升用户满意度并保持计算效率。

摘要:

交互式AI系统,如推荐引擎和虚拟助手,通常使用静态用户资料和预定义规则来个性化交互。然而,这些方法往往无法捕捉用户偏好和情境的动态特性。本研究提出了一种理论框架和实际实现方法,用于将持续反馈循环整合到个性化算法中,以实现实时适应。通过持续收集和分析用户反馈,AI系统可以动态调整其推荐、响应和交互,以更好地与用户当前的情境和偏好对齐。我们为我们的自适应个性化算法的收敛性和遗憾界提供了理论保证。我们在三个领域——推荐系统、虚拟助手和自适应学习平台——的实验评估表明,动态个性化相比静态方法将用户满意度提高了15-23%,同时保持计算效率。我们研究了持续反馈机制的实现挑战,评估了其对用户体验和满意度的影响,并提供了个性化质量、计算开销和用户疲劳之间的权衡的全面分析。

英文摘要:

Interactive AI systems, such as recommendation engines and virtual assistants, commonly use static user profiles and predefined rules to personalize interactions. However, these methods often fail to capture the dynamic nature of user preferences and context. This study proposes a theoretical framework and practical implementation for integrating continuous feedback loops into personalization algorithms to enable real-time adaptation. By continuously collecting and analyzing user feedback, the AI system can dynamically adjust its recommendations, responses, and interactions to better align with the user's current context and preferences. We provide theoretical guarantees for the convergence and regret bounds of our adaptive personalization algorithm. Our experimental evaluation across three domains-recommendation systems, virtual assistants, and adaptive learning platforms-demonstrates that dynamic personalization improves user satisfaction by 15-23% compared to static methods while maintaining computational efficiency. We investigated the implementation challenges of continuous feedback mechanisms, evaluated their impact on user experience and satisfaction, and provided a comprehensive analysis of the trade-offs between personalization quality, computational overhead, and user fatigue.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑