Start Date
10-12-2017 12:00 AM
Description
Context-Aware Recommender Systems (CARSs) are becoming commonplace. Yet, there is a paucity of studies that investigates how such systems could affect usage behavior from a user-system interaction perspective. Building on the Social Interdependence Theory (SIT), we construct a research model that posits cooperative learning as a trait of users’ interactions with CARSs and outline a proposed empirical study for validating the hypothesized relationships in this model. Specifically, we draw on interdependencies in human-machine relationships to postulate positive interdependence as an antecedent of users’ promotive interaction with CARSs, which in turn, dictates the performance of such recommender systems. Furthermore, we introduce scrutability features as design interventions that can be harnessed by developers to mitigate the impact of users’ promotive interaction on the performance of CARSs.
Recommended Citation
Jiang, Na; Tan, Chee-Wee; WANG, Weiquan; Liu, Hefu; and Gu, Jibao, "Understanding Cooperative Learning in Context-Aware Recommender Systems: A User-System Interaction Perspective" (2017). ICIS 2017 Proceedings. 11.
https://aisel.aisnet.org/icis2017/HCI/Presentations/11
Understanding Cooperative Learning in Context-Aware Recommender Systems: A User-System Interaction Perspective
Context-Aware Recommender Systems (CARSs) are becoming commonplace. Yet, there is a paucity of studies that investigates how such systems could affect usage behavior from a user-system interaction perspective. Building on the Social Interdependence Theory (SIT), we construct a research model that posits cooperative learning as a trait of users’ interactions with CARSs and outline a proposed empirical study for validating the hypothesized relationships in this model. Specifically, we draw on interdependencies in human-machine relationships to postulate positive interdependence as an antecedent of users’ promotive interaction with CARSs, which in turn, dictates the performance of such recommender systems. Furthermore, we introduce scrutability features as design interventions that can be harnessed by developers to mitigate the impact of users’ promotive interaction on the performance of CARSs.