Advances in Information Retrieval

The field of information retrieval is moving towards a deeper understanding of the impact of large language models (LLMs) on retrieval systems. Researchers are investigating how LLM-generated content affects term-based retrieval models and exploring methods to diversify queries to reflect the diversity of search engine users. There is also a growing interest in evaluating and improving patent novelty search systems, as well as developing more effective semantic search models. Noteworthy papers include: Research on Evaluation Methods for Patent Novelty Search Systems, which proposes a comprehensive evaluation methodology for patent novelty search systems. LexSemBridge: Fine-Grained Dense Representation Enhancement, which introduces a unified framework to enhance dense query representations for fine-grained retrieval tasks.

Sources

How Do LLM-Generated Texts Impact Term-Based Retrieval Models?

Demographically-Inspired Query Variants Using an LLM

Research on Evaluation Methods for Patent Novelty Search Systems and Empirical Analysis

Semantic Search for Information Retrieval

LexSemBridge: Fine-Grained Dense Representation Enhancement through Token-Aware Embedding Augmentation

Built with on top of