Learning User Preferences to Incentivize Exploration in the Sharing Economy

Authors: Christoph Hirnschall, Adish Singla, Sebastian Tschiatschek, Andreas Krause

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We provide formal guarantees on the performance of our algorithm and test the viability of our approach in a user study with data of apartments on Airbnb. Our findings suggest that our approach is well-suited to learn appropriate incentives and increase exploration on the investigated platform.
Researcher Affiliation Collaboration Christoph Hirnschall ETH Zurich Zurich, Switzerland chirnsch@gmail.com Adish Singla MPI-SWS Saarbr ucken, Germany adishs@mpi-sws.org Sebastian Tschiatschek Microsoft Research* Cambridge, United Kingdom setschia@microsoft.com Andreas Krause ETH Zurich Zurich, Switzerland krausea@ethz.ch *Work performed while at ETH Zurich.
Pseudocode Yes Algorithm 1: OL Online Learning; Algorithm 2: Co OL Coordinated Online Learning; Function 3: AProj Approximate Projection
Open Source Code No The paper does not contain any explicit statements about releasing source code for the methodology or provide a link to a code repository.
Open Datasets Yes Airbnb dataset. Using data of Airbnb apartments from insideairbnb.com, we created a dataset of 20 apartments as follows: we chose apartments from 4 types in New York City by location (Manhattan or Brooklyn) and number of reviews (high, 20 or low, 2). (1Data from insideairbnb.com.)
Dataset Splits No The paper does not specify explicit training, validation, and test dataset splits (e.g., percentages, sample counts, or predefined split references).
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running the experiments.
Software Dependencies No The paper does not provide specific software names with version numbers (e.g., libraries, frameworks, or solvers) that would be needed to replicate the experiment.
Experiment Setup Yes Utility gain. The utility gain u for getting a review for infrequently reviewed apartments is set to u = 40 in our experiments, based on referral discounts given by Airbnb in the past. Loss function. ...we use a delta value of 20. Structure. Due to the small number of apartments, we consider a non-contextual setting with d = 1 and use an rbounded hemimetric structure to model the relationship of the tasks, where r is set to 40 to avoid recommending incentives pt > u.