In computing, the aim of personalization is to train a model that caters to a specific individual or group of people by optimizing one or more performance metrics and adhering to specific constraints. In this paper, we discuss the need for personalization in affective and personality computing (hereinafter referred to as affective computing). We present a survey of state-of-the-art approaches for personalization in affective computing. Our review spans training techniques and objectives towards the personalization of affective computing models. We group existing approaches into seven categories: (1) Target-specific Models, (2) Group-specific Models, (3) Weighting-based Approaches, (4) Fine-tuning Approaches, (5) Multitask Learning, (6) Generative-based Models, and (7) Feature Augmentation. Additionally, we provide a statistical meta-analysis of the surveyed literature, analyzing the prevalence of different affective computing tasks, interaction modes, interaction contexts, and the level of personalization among the surveyed works. Based on that, we provide a road-map for those who are interested in exploring this direction.

翻译：在计算机领域，个性化的目标是通过针对特定个体或群体优化一项或多项性能指标并满足特定约束，来训练适配其需求的模型。本文探讨了情感与个性计算（以下简称情感计算）中个性化的必要性，系统综述了情感计算个性化方法的最新研究进展。我们从训练技术和优化目标两个维度梳理了情感计算模型的个性化实现路径，将现有方法归纳为七类：（1）目标特定模型、（2）群体特定模型、（3）加权方法、（4）微调方法、（5）多任务学习、（6）生成模型及（7）特征增强方法。此外，我们对所综述文献进行了统计元分析，揭示了不同情感计算任务、交互模式、交互情境及个性化程度在现有研究中的分布特征。基于此，本文为有意探索该方向的研究者提供了路线图。