Explainable AI (XAI) is widely viewed as a sine qua non for ever-expanding AI
research. A better understanding of the needs of XAI users, as well as
human-centered evaluations of explainable models are both a necessity and a
challenge. In this paper, we explore how HCI and AI researchers conduct user
studies in XAI applications based on a systematic literature review. After
identifying and thoroughly analyzing 85 core papers with human-based XAI
evaluations over the past five years, we categorize them along the measured
characteristics of explanatory methods, namely trust, understanding, fairness,
usability, and human-AI team performance. Our research shows that XAI is
spreading more rapidly in certain application domains, such as recommender
systems than in others, but that user evaluations are still rather sparse and
incorporate hardly any insights from cognitive or social sciences. Based on a
comprehensive discussion of best practices, i.e., common models, design
choices, and measures in user studies, we propose practical guidelines on
designing and conducting user studies for XAI researchers and practitioners.
Lastly, this survey also highlights several open research directions,
particularly linking psychological science and human-centered XAI.

本文通过系统文献综述，评估了 85 篇人基于 XAI 评估的核心论文，并对解释方法的可信、可懂、公平、可用性和人工智能团队表现等进行了归类。作者提出了关于设计和进行 XAI 用户研究的最佳实践指南，并强调了将心理科学和以人为中心的 XAI 联系起来的几个开放研究方向。

面向人类中心的可解释人工智能：模型解释的用户研究

Towards Human-centered Explainable AI: User Studies for Model Explanations

Deep learning has produced state-of-the-art results for a variety of tasks.
While such approaches for supervised learning have performed well, they assume
that training and testing data are drawn from the same distribution, which may
not always be the case. As a complement to this challenge, single-source
unsupervised domain adaptation can handle situations where a network is trained
on labeled data from a source domain and unlabeled data from a related but
different target domain with the goal of performing well at test-time on the
target domain. Many single-source and typically homogeneous unsupervised deep
domain adaptation approaches have thus been developed, combining the powerful,
hierarchical representations from deep learning with domain adaptation to
reduce reliance on potentially-costly target data labels. This survey will
compare these approaches by examining alternative methods, the unique and
common elements, results, and theoretical insights. We follow this with a look
at application areas and open research directions.

该研究论文对深度学习与非监督领域自适应进行了综述并对其中一些方法进行了比较，同时探讨了应用领域及未来研究方向。

无监督深度域适应综述

A Survey of Unsupervised Deep Domain Adaptation

Device-to-device (D2D) communication underlaying cellular networks allows
mobile devices such as smartphones and tablets to use the licensed spectrum
allocated to cellular services for direct peer-to-peer transmission. D2D
communication can use either one-hop transmission (i.e., in D2D direct
communication) or multi-hop cluster-based transmission (i.e., in D2D local area
networks). The D2D devices can compete or cooperate with each other to reuse
the radio resources in D2D networks. Therefore, resource allocation and access
for D2D communication can be treated as games. The theories behind these games
provide a variety of mathematical tools to effectively model and analyze the
individual or group behaviors of D2D users. In addition, game models can
provide distributed solutions to the resource allocation problems for D2D
communication. The aim of this article is to demonstrate the applications of
game-theoretic models to study the radio resource allocation issues in D2D
communication. The article also outlines several key open research directions.

本文章展示了博弈论模型在 D2D 通信资源分配问题中的应用，并概述了一些重要的开放性研究方向。