24–26 Mar 2025 In-Person Event
Helmholtz-Zentrum Berlin für Materialien und Energie
Europe/Berlin timezone

Leveraging Large Language Models for Automated Outreach in DAPHNE4NFDI TA4

Not scheduled
20m
Helmholtz-Zentrum Berlin für Materialien und Energie

Helmholtz-Zentrum Berlin für Materialien und Energie

Campus Adlershof Albert-Einstein-Straße 15 12489 Berlin
Poster Poster

Speaker

Dr Öezgül Öeztürk (University of Siegen)

Description

One of the main goals of TA4 is effective science communication and outreach for maximizing the impact of DAPHNE4NFDI. For managing the outreach activities with less time and sources we explore the integration of Large Language Models (LLMs) to automate key aspects of outreach in TA4 at DAPHNE4NFDI.
We propose an LLM-driven workflow for automatization of (1) event summarization, generating structured reports, social media posts, and news from workshops, meetings and conferences; and (2) social media monitoring and insights, employing AI-based analytics to optimize content strategies and track audience engagement.
Preliminary tests demonstrate that LLMs can generate targeted outreach messages, and provide data-driven insights into audience behavior. This approach serves as a model for integrating LLMs into research outreach, supporting more efficient science communication.

Primary author

Dr Öezgül Öeztürk (University of Siegen)

Co-authors

Mr Agha Mohammad Raza (University of Siegen) Dr Aliaksandr Leonau (University of Siegen) Dr Amir Tosson (University of Siegen) Ms Michelle Dargasz (University of Siegen)

Presentation materials

There are no materials yet.