Who is the Pronoun Pro? Comparing LLMs and Trained Human Annotators in Pronoun Resolution

Key Points

• The research aims to compare the performance of large language models and trained human annotators in resolving pronouns in text.
• Conducted a comparison between large language models and human expert annotators.
• Evaluated the accuracy of pronoun resolution in various linguistic contexts.
• Utilized quantitative metrics to assess performance differences.
• Large language models showed comparable accuracy to human annotators in pronoun resolution tasks.
• Efficiency of language models is highlighted, potentially reducing the need for extensive human input.