NovAScore: A New Automated Metric for Evaluating Document Level Novelty

TL;DR

NovAScore is an automated, interpretable metric designed to evaluate document-level novelty, addressing the limitations of human annotation and correlating well with human judgments.

Contribution

The paper introduces NovAScore, a novel automated metric for evaluating document novelty that combines atomic information scores with dynamic weighting for improved interpretability and flexibility.

Findings

Achieves 0.626 correlation with human judgments on TAP-DLND 1.0 dataset.

Achieves 0.920 Pearson correlation on internal human-annotated dataset.

Provides detailed analysis of a document's novelty and importance.

Abstract

The rapid expansion of online content has intensified the issue of information redundancy, underscoring the need for solutions that can identify genuinely new information. Despite this challenge, the research community has seen a decline in focus on novelty detection, particularly with the rise of large language models (LLMs). Additionally, previous approaches have relied heavily on human annotation, which is time-consuming, costly, and particularly challenging when annotators must compare a target document against a vast number of historical documents. In this work, we introduce NovAScore (Novelty Evaluation in Atomicity Score), an automated metric for evaluating document-level novelty. NovAScore aggregates the novelty and salience scores of atomic information, providing high interpretability and a detailed analysis of a document's novelty. With its dynamic weight adjustment scheme, NovAScore offers enhanced flexibility and an additional dimension to assess both the novelty level and the importance of information within a document. Our experiments show that NovAScore strongly correlates with human judgments of novelty, achieving a 0.626 Point-Biserial correlation on the TAP-DLND 1.0 dataset and a 0.920 Pearson correlation on an internal human-annotated dataset.