AlignScore: Evaluating Factual Consistency with a Unified Alignment Function
This addresses the problem of assessing diverse factual inconsistencies in text generation for applications requiring reliable evaluation, representing a strong specific gain rather than a foundational breakthrough.
The paper tackles the challenge of automatically evaluating factual consistency in text generation by proposing AlignScore, a holistic metric based on a unified alignment function trained on 4.7M examples from diverse tasks, which achieves substantial improvements over previous metrics and matches or outperforms larger models like ChatGPT and GPT-4 on 22 benchmarks.
Many text generation applications require the generated text to be factually consistent with input information. Automatic evaluation of factual consistency is challenging. Previous work has developed various metrics that often depend on specific functions, such as natural language inference (NLI) or question answering (QA), trained on limited data. Those metrics thus can hardly assess diverse factual inconsistencies (e.g., contradictions, hallucinations) that occur in varying inputs/outputs (e.g., sentences, documents) from different tasks. In this paper, we propose AlignScore, a new holistic metric that applies to a variety of factual inconsistency scenarios as above. AlignScore is based on a general function of information alignment between two arbitrary text pieces. Crucially, we develop a unified training framework of the alignment function by integrating a large diversity of data sources, resulting in 4.7M training examples from 7 well-established tasks (NLI, QA, paraphrasing, fact verification, information retrieval, semantic similarity, and summarization). We conduct extensive experiments on large-scale benchmarks including 22 evaluation datasets, where 19 of the datasets were never seen in the alignment training. AlignScore achieves substantial improvement over a wide range of previous metrics. Moreover, AlignScore (355M parameters) matches or even outperforms metrics based on ChatGPT and GPT-4 that are orders of magnitude larger.