A study on user acceptance of proactive in-vehicle recommender systems

Abstract

Modern in-vehicle information systems (IVIS) are able to provide a large amount of data to the driver. If every information which might be of interest is delivered directly to the driver, information overload becomes a serious problem. Recommender systems are a promising approach to reduce information overload but they are mainly designed for desktop systems or mobile devices. In-vehicle recommender systems have to cope with interaction restrictions and limited cognitive resources of the driver. Therefore, we investigate proactive recommender systems, where recommendations are pushed automatically. The contribution of this paper is a user study in a real world setup to investigate the acceptance of a proactive recommender system while driving. The evaluation is based on the Technology Acceptance Model (TAM). As perceived ease of use is crucial for acceptance, we design an in-vehicle user interface for proactive recommendations. Our results show that our proactive recommender is perceived as helpful and assisting and is not obtrusive and distracting while driving. We also found that clear information delivery and trust is crucial for the acceptance of in-vehicle recommendations.