Human Interaction with Recommendation Systems

By Sven Schmit, Carlos Riquelme. 2017

53 citations recorded by the source corpus.

Many recommendation algorithms rely on user data to generate recommendations. However, these recommendations also affect the data obtained from future users. This work aims to understand the effects of this dynamic interaction. We propose a simple model where users with heterogeneous preferences arrive over time. Based on this model, we prove that naive estimators, i.e. those which ignore this feedback loop, are not consistent. We show that consistent estimators are efficient in the presence of myopic agents. Our results are validated using extensive simulations.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Introduction to Multi-Armed Bandits (2019, background): “ns with a constant probability, as long as reward 0 can be realized), and arm 2 is never tried again. A similar but much stronger negative guarantee is stated in Exercise 11.2. A recent line of work (Schmit and Riquelme, 2018; Kannan et al., 2018; Bastani et al., 2018; Raghavan et al., 2018). ﬁnds t”
Data Feedback Loops: Model-driven Amplification of Dataset Biases (2022, result): “Our work is also closely aligned with the study of feedback loops in recommendation systems [58, 55].”