Validating Simulations of User Query Variants

TL;DR

This paper evaluates the realism of simulated user query variants in information retrieval, demonstrating their potential to mimic real user queries and proposing a new method that improves simulation quality based on TREC data.

Contribution

It introduces a simple, effective method for simulating user query variants that better reproduces real queries compared to existing approaches.

Findings

Simulated queries closely match real queries in retrieval effectiveness.

The new method improves query term similarity and topic score distribution reproduction.

Simulations are useful for system evaluation when real user data is unavailable.

Abstract

System-oriented IR evaluations are limited to rather abstract understandings of real user behavior. As a solution, simulating user interactions provides a cost-efficient way to support system-oriented experiments with more realistic directives when no interaction logs are available. While there are several user models for simulated clicks or result list interactions, very few attempts have been made towards query simulations, and it has not been investigated if these can reproduce properties of real queries. In this work, we validate simulated user query variants with the help of TREC test collections in reference to real user queries that were made for the corresponding topics. Besides, we introduce a simple yet effective method that gives better reproductions of real queries than the established methods. Our evaluation framework validates the simulations regarding the retrieval performance, reproducibility of topic score distributions, shared task utility, effort and effect, and query term similarity when compared with real user query variants. While the retrieval effectiveness and statistical properties of the topic score distributions as well as economic aspects are close to that of real queries, it is still challenging to simulate exact term matches and later query reformulations.