WP3 invites to a digital seminar on AI for local journalism. Speaker is the Postdoctoral Researcher Dr Reshmi G. Pillai from Vrije University Amsterdam.

Journalism has been transformed by technological advancements over the last two decades, causing significant economic disruption among other challenges. The emergence of Artificial Intelligence has accelerated this paradigm shift, presenting tremendous opportunities and equally profound challenges to news organizations worldwide. This transition poses a particular concern to regional and local media organizations, as they are often at increased risk of being at the disadvantaged end of the technological knowledge and adoption. Local journalism, however, remains a vital component of democratic societies, by voicing the issues and concerns from local communities, and contextualizing global/national news stories, thus engaging news audience better with collective discourses.

Our research investigates, in collaboration with a regional broadcaster in the Netherlands, how AI can best enhance journalistic processes, while maintaining the agency with the journalists. We conducted design thinking sprints with the journalists of the broadcaster, starting by identifying the challenges within the news-making process, and moving towards clearly defined problem statements and solution approaches. The talk also presents our further research motivated by these findings from the design thinking sprints. We outline a prototype to build a Retrieval Augmented Generation (RAG) system from information sources such as past news reports, to provide background information for an incoming news story. Also, addressing the problem of news content and headline optimization, we analyse the textual features, on a dataset of ~500k tweets, associated with audience engagement, for national and regional news media within the Netherlands, as pointers for generating engaging content.

This research is part of the consortium project ‘Towards Responsible AI for Local Journalism’ (link here) with researchers from Vrije University Amsterdam, University of Amsterdam, University of Groningen (the Netherlands), LMU Munich (Germany), and University of Stavanger (Norway).

Speaker: Dr Reshmi G. Pillai

Ms. Pillai focuses on the adoption and implementation of artificial intelligence techniques, specifically natural language processing in the context of local journalism. Her research interests are in analyzing text using natural language processing methods for a better understanding of behaviors and states of individuals and the society. Previously, she was a Lecturer at the Informatics Institute, University of Amsterdam, assisting in the Master’s program in Information Studies (Data Science track). She holds a PhD from the University of Wolverhampton for her thesis on the expressions of psychological stress in tweets.