手法を比較

	幻覚検出	BERT埋め込み
分野	テキストマイニング	テキストマイニング
系統	Process / pipeline	Process / pipeline
提唱年	2020 (faithfulness framing); 2023 (SelfCheckGPT)	2019
提唱者	Established as a formal task by Maynez et al. (2020); SelfCheckGPT zero-resource variant by Manakul et al. (2023)	Devlin, Chang, Lee & Toutanova (Google AI)
種類	NLP evaluation / quality-assurance pipeline	Contextual transformer text-representation method
原典	Maynez, J., Narayan, S., Bohnet, B., & McDonald, R. (2020). On Faithfulness and Factuality in Abstractive Summarization. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL), 1906-1919.	Devlin, J., Chang, M.-W., Lee, K. & Toutanova, K. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. NAACL-HLT, 4171-4186.
別名	factual consistency checking, faithfulness evaluation, LLM output verification, Hallüsinasyon Tespiti (Factual Consistency)	contextual embeddings, transformer embeddings, BERT Tabanlı Metin Gömülmeleri
関連	5	4
概要	Hallucination detection is a natural-language-processing pipeline that measures whether the output of a language model is consistent with a reference source document or with verifiable facts. Formalised as a faithfulness evaluation task by Maynez et al. (2020) and extended to a zero-resource black-box setting by Manakul et al. (2023) with SelfCheckGPT, the approach is used to flag unreliable LLM outputs in high-stakes domains such as medicine, law, and journalism.	BERT-based text embeddings, introduced by Devlin and colleagues at Google AI in 2019, turn text into context-sensitive dense vectors using a bidirectional Transformer encoder. Because the meaning of a word shifts with its context, BERT produces richer representations than static methods such as Word2Vec or topic models like LDA.