Results of the summarizing process for the arXiv paper: 2202.03629v1

Natural Language Generation (NLG) has made significant advancements in recent years due to the development of deep learning technologies such as Transformer-based language models. This progress has enabled more fluent and coherent natural language generation for downstream tasks like abstractive summarization, dialogue generation, and data-to-text generation. However, NLG also generates hallucinated texts that fail to meet users' expectations in real-world scenarios. To address this issue, researchers have presented various evaluation and mitigation methods for hallucinations in different tasks. In this survey paper, Ziwei Ji and colleagues from Hong Kong University of Science and Technology provide an overview of research progress and challenges related to the hallucination problem in NLG. The authors categorize hallucinations into intrinsic (generated text contradicts factual information or common sense) and extrinsic (generated text does not relate to the input). They discuss contributors to hallucinations such as data quality issues or errors during training or inference. Statistical metrics based on perplexity scores or n-gram overlap ratios as well as model-based metrics like adversarial attacks or gradient-based saliency maps are proposed for evaluating hallucination severity. Human evaluation methods like crowd-sourcing are also discussed. Various mitigation methods including data augmentation or filtering and regularization or fine-tuning are presented for reducing hallucinations in NLG systems. The authors emphasize the need for a standardized evaluation framework for future directions in metrics and mitigation methods. In summary, this survey paper provides a comprehensive overview of research progress related to the hallucination problem in NLG with proposed metrics and mitigation methods that could facilitate collaborative efforts among researchers working on different downstream tasks.

• - Natural Language Generation (NLG) has made significant advancements in recent years due to the development of deep learning technologies such as Transformer-based language models.
• - NLG enables more fluent and coherent natural language generation for downstream tasks like abstractive summarization, dialogue generation, and data-to-text generation.
• - However, NLG also generates hallucinated texts that fail to meet users' expectations in real-world scenarios.
• - Researchers have presented various evaluation and mitigation methods for hallucinations in different tasks.
• - Ziwei Ji and colleagues from Hong Kong University of Science and Technology provide an overview of research progress and challenges related to the hallucination problem in NLG.
• - The authors categorize hallucinations into intrinsic (generated text contradicts factual information or common sense) and extrinsic (generated text does not relate to the input).
• - They discuss contributors to hallucinations such as data quality issues or errors during training or inference.
• - Statistical metrics based on perplexity scores or n-gram overlap ratios as well as model-based metrics like adversarial attacks or gradient-based saliency maps are proposed for evaluating hallucination severity.
• - Human evaluation methods like crowd-sourcing are also discussed.
• - Various mitigation methods including data augmentation or filtering and regularization or fine-tuning are presented for reducing hallucinations in NLG systems.
• - The authors emphasize the need for a standardized evaluation framework for future directions in metrics and mitigation methods.

Summary: NLG is a technology that helps computers generate sentences like humans. It has gotten better because of new learning methods. NLG can help with things like making summaries or talking to people. But sometimes, the computer makes mistakes and says things that don't make sense. Scientists are working on ways to fix these mistakes. Definitions- Natural Language Generation (NLG): Technology that helps computers generate sentences like humans. - Deep learning technologies: A type of machine learning where the computer learns from examples. - Transformer-based language models: A specific type of deep learning model used for natural language processing tasks. - Abstractive summarization: Creating a summary by understanding the meaning of text and generating new sentences. - Dialogue generation: Generating conversation between a computer and a person. - Data-to-text generation: Generating written text from data or information. - Hallucinated texts: Sentences generated by NLG that do not make sense or are incorrect. - Evaluation methods: Ways to measure how well NLG systems are working. - Mitigation methods: Ways to reduce errors in NLG systems. - Standardized evaluation framework: A set of guidelines for how to test and compare different NLG systems.

Natural Language Generation (NLG): An Overview of Hallucination Problem and Mitigation Methods

Recent advancements in deep learning technologies such as Transformer-based language models have enabled more fluent and coherent natural language generation (NLG) for downstream tasks like abstractive summarization, dialogue generation, and data-to-text generation. However, NLG also generates hallucinated texts that fail to meet users' expectations in real-world scenarios. To address this issue, researchers have presented various evaluation and mitigation methods for hallucinations in different tasks. In this survey paper, Ziwei Ji and colleagues from Hong Kong University of Science and Technology provide an overview of research progress and challenges related to the hallucination problem in NLG. The authors categorize hallucinations into intrinsic (generated text contradicts factual information or common sense) and extrinsic (generated text does not relate to the input). They discuss contributors to hallucinations such as data quality issues or errors during training or inference. This survey paper provides a comprehensive overview of research progress related to the hallucination problem in NLG with proposed metrics and mitigation methods that could facilitate collaborative efforts among researchers working on different downstream tasks.

Evaluation Metrics

The authors propose several statistical metrics based on perplexity scores or n-gram overlap ratios as well as model-based metrics like adversarial attacks or gradient-based saliency maps for evaluating hallucination severity. Human evaluation methods like crowd-sourcing are also discussed.

Mitigation Methods

Various mitigation methods including data augmentation or filtering and regularization or fine-tuning are presented for reducing hallucinations in NLG systems. The authors emphasize the need for a standardized evaluation framework for future directions in metrics and mitigation methods. In summary, this survey paper provides a comprehensive overview of research progress related to the hallucination problem in NLG with proposed metrics and mitigation methods that could facilitate collaborative efforts among researchers working on different downstream tasks