Cocktail: A Comprehensive Information Retrieval Benchmark with LLM-Generated Documents Integration

TL;DR

Cocktail is a new comprehensive benchmark designed to evaluate information retrieval models in a landscape increasingly populated by both human-written and LLM-generated content, addressing a critical gap in IR research tools.

Contribution

We introduce Cocktail, a diverse IR benchmark with mixed datasets and a new dataset to evaluate models in the LLM era, along with extensive experiments revealing performance and bias trade-offs.

Findings

Neural retrieval models show a trade-off between ranking performance and source bias.

The benchmark highlights the need for balanced IR system design in the presence of LLM-generated content.

Cocktail provides a standardized platform for future IR research in the context of LLMs.

Abstract

The proliferation of Large Language Models (LLMs) has led to an influx of AI-generated content (AIGC) on the internet, transforming the corpus of Information Retrieval (IR) systems from solely human-written to a coexistence with LLM-generated content. The impact of this surge in AIGC on IR systems remains an open question, with the primary challenge being the lack of a dedicated benchmark for researchers. In this paper, we introduce Cocktail, a comprehensive benchmark tailored for evaluating IR models in this mixed-sourced data landscape of the LLM era. Cocktail consists of 16 diverse datasets with mixed human-written and LLM-generated corpora across various text retrieval tasks and domains. Additionally, to avoid the potential bias from previously included dataset information in LLMs, we also introduce an up-to-date dataset, named NQ-UTD, with queries derived from recent events. Through conducting over 1,000 experiments to assess state-of-the-art retrieval models against the benchmarked datasets in Cocktail, we uncover a clear trade-off between ranking performance and source bias in neural retrieval models, highlighting the necessity for a balanced approach in designing future IR systems. We hope Cocktail can serve as a foundational resource for IR research in the LLM era, with all data and code publicly available at \url{https://github.com/KID-22/Cocktail}.