Authors: Yining Juan, Chung-Chi Chen, Hen-Hsen Huang, Hsin-Hsi Chen

Published on: September 27, 2024

Impact Score: 7.6

Arxiv code: Arxiv:2409.18677

Summary

• What is new: Our paper introduces the multi-question generation (MQG) task specifically designed for earnings call contexts.
• Why this is important: Traditional methods for anticipating audience questions during professional events are inefficient and imprecise, especially for large or diverse groups.
• What the research proposes: We propose a novel annotation technique and a retriever-enhanced strategy to extract relevant information from earnings call transcripts to generate potential questions.
• Results: Our approach achieves notable excellence in accuracy, consistency, and perplexity of the generated questions.

Technical Details

Technological frameworks used: NLP techniques

Models used: Retriever-enhanced strategy

Data used: Earnings call transcripts

Potential Impact

Financial analysis firms, corporate communication departments, and companies conducting earnings calls could benefit from these insights.

Want to implement this idea in a business?

We have generated a startup concept here: InquisitiveAI.