When the machine learns from users, is it helping or snooping?

作者：

Highlights：

• When a system is transparent about its learning (machine learning cue), users consider the system as a helper.

• When users perceive an AI system to be a helper, their frustration is low and they express more trust in the system.

• Machine learning cue is effective regardless of the system's performance and explicitness of the cue.

• No evidence to suggest that machine learning cue evokes privacy concerns.

摘要

•When a system is transparent about its learning (machine learning cue), users consider the system as a helper.•When users perceive an AI system to be a helper, their frustration is low and they express more trust in the system.•Machine learning cue is effective regardless of the system's performance and explicitness of the cue.•No evidence to suggest that machine learning cue evokes privacy concerns.

论文关键词：Algorithm,Machine learning cue,HAII-TIME,Helper heuristic,System performance,Perceived frustration,Privacy concern,Trust

论文评审过程：Received 21 April 2022, Revised 27 July 2022, Accepted 30 July 2022, Available online 3 August 2022, Version of Record 18 September 2022.

论文官网地址：https://doi.org/10.1016/j.chb.2022.107427