A Sensitivity Analysis of the MSMARCO Passage Collection

TL;DR

This study investigates how the MSMARCO passage collection's system rankings are affected by assuming more passages as relevant, finding that rankings are stable despite score variations when relevance criteria are expanded.

Contribution

It provides a sensitivity analysis of MSMARCO's relevance annotations, demonstrating the robustness of system rankings against changes in relevance assumptions.

Findings

Run scores vary with additional relevant passages.

System rankings remain relatively stable despite score fluctuations.

Supports the original relevance annotation methodology.

Abstract

The recent MSMARCO passage retrieval collection has allowed researchers to develop highly tuned retrieval systems. One aspect of this data set that makes it distinctive compared to traditional corpora is that most of the topics only have a single answer passage marked relevant. Here we carry out a "what if" sensitivity study, asking whether a set of systems would still have the same relative performance if more passages per topic were deemed to be "relevant", exploring several mechanisms for identifying sets of passages to be so categorized. Our results show that, in general, while run scores can vary markedly if additional plausible passages are presumed to be relevant, the derived system ordering is relatively insensitive to additional relevance, providing support for the methodology that was used at the time the MSMARCO passage collection was created.