paper-with-me

홈 › Papers

Using Similarity to Evaluate Factual Consistency in Summaries

2024-09-23 · Yuxuan Ye, Edwin Simpson, Raul Santos Rodriguez

Cutting-edge abstractive summarisers generate fluent summaries, but the factuality of the generated text is not guaranteed. Early summary factuality evaluation metrics are usually based on n-gram overlap and embedding similarity, but are reported fail to align with human annotations. Therefore, many techniques for detecting factual inconsistencies build pipelines around natural language inference (NLI) or question-answering (QA) models with additional supervised learning steps. In this paper, we revisit similarity-based metrics, showing that this failure stems from the comparison text selection and its granularity. We propose a new zero-shot factuality evaluation metric, Sentence-BERT Score (SBERTScore), which compares sentences between the summary and the source document. It outperforms widely-used word-word metrics including BERTScore and can compete with existing NLI and QA-based factuality metrics on the benchmark without needing any fine-tuning. Our experiments indicate that each technique has different strengths, with SBERTScore particularly effective in identifying correct summaries. We demonstrate how a combination of techniques is more effective in detecting various types of error.

📄 PDF Abstract BibTeX arXiv:2409.15090

Code (0)

등록된 구현이 없습니다.

Tasks

Natural Language InferenceQuestion AnsweringSentence

Methods 이 논문이 사용한 방법론

ALIGN In the ALIGN method, visual and language representations are jointly trained from noisy image alt-text data. The image and text encoders are learned via contrastive loss…

Similar Papers 제목 키워드 기반

Evaluating the Factual Consistency of Large Language Models Through News Summarization

2022-11-15 · Derek Tam, Anisha Mascarenhas, Shiyue Zhang, Sarah Kwan 외

While large language models (LLMs) have proven to be effective on a large variety of tasks, they are also known to hallucinate information. To measure whether an LLM prefers factually consistent continuations of its inpu…

News Summarization

ReFEree: Reference-Free and Fine-Grained Method for Evaluating Factual Consistency in Real-World Code Summarization

2026-04-12 · Suyoung Bae, CheolWon Na, Jaehoon Lee, Yumin Lee 외 arxiv

As Large Language Models (LLMs) have become capable of generating long and descriptive code summaries, accurate and reliable evaluation of factual consistency has become a critical challenge. However, previous evaluation…

Enhancing Factual Consistency of Abstractive Summarization

2020-03-19 · NAACL 2021 4 · Chenguang Zhu, William Hinthorn, Ruochen Xu, Qingkai Zeng 외

Automatic abstractive summaries are found to often distort or fabricate facts in the article. This inconsistency between summary and original text has seriously impacted its applicability. We propose a fact-aware summari…

Abstractive Text SummarizationGraph Attention

Masked Summarization to Generate Factually Inconsistent Summaries for Improved Factual Consistency Checking

2022-05-04 · Findings (NAACL) 2022 7 · Hwanhee Lee, Kang Min Yoo, Joonsuk Park, Hwaran Lee 외

Despite the recent advances in abstractive summarization systems, it is still difficult to determine whether a generated summary is factual consistent with the source text. To this end, the latest approach is to train a …

Abstractive Text Summarization

Exploring the Factual Consistency in Dialogue Comprehension of Large Language Models

2023-11-13 · Shuaijie She, ShuJian Huang, Xingyun Wang, Yanke Zhou 외

LLMs (Large Language Models) usually interact with users in the form of dialogue and generate responses following their instructions, which naturally require dialogue comprehension abilities. However, dialogue comprehens…