Succes and Failure Factors of Generative AI in a Chat Application of Dutch Railways
This program is tentative and subject to change.
Since the introduction of ChatGPT in 2020, there has been exponential growth in the application of Generative AI (GAI) in the business processes of public transport organizations. However, organizations still struggle with the implementation of such applications and want to gain better insight into the factors that positively and negatively influence their introduction and use in practice. To investigate this, we conducted an exploratory case study with 17 in-depth interviews and an online survey among 59 users and GAI specialists of the ChatNS application at the Dutch Railways(NS). In our study we found five factors that according to the analyzed literature and the interviewed GAI users influence the use of GAI in a practical setting. As a tool for future research we created a maturity framework with five KPIs linked to Success and Failure Factors (SFFs) for future GAI applications.
This program is tentative and subject to change.
Thu 2 OctDisplayed time zone: Hawaii change
11:30 - 12:40 | Generative AI in Software EngineeringESEM - Industry, Government, and Community Track / ESEM - Technical Track / ESEM - Registered Reports Track at Kaiulani II Chair(s): Amiangshu Bosu Wayne State University | ||
11:30 17mTalk | Succes and Failure Factors of Generative AI in a Chat Application of Dutch Railways ESEM - Industry, Government, and Community Track Elise Peusen Utrecht University of Applied Sciences, Leo van der Meulen NS, Hennie Huijgens Utrecht University of Applied Sciences, Lucque Schmeitz Utrecht University of Applied Sciences | ||
11:47 17mTalk | Evaluating Generative AI Tools for Personalised Offline Recommendations: A Comparative Study ESEM - Registered Reports Track Rafael Salinas Universidad de Cuenca, Otto Parra Universidad de Cuenca, Condori-Fernandez Nelly Universidad de Santiago de Compostela, Maria Fernanda Granda Juca Universidad de Cuenca | ||
12:05 17mTalk | Using Biometrics to Understand AI-Assisted Coding Performance and its Perception: a Registered Report ESEM - Registered Reports Track Nadja Brix Koch IT University of Copenhagen, Theis Helth Stensgaard IT University of Copenhagen, Paolo Tell IT University of Copenhagen, Denmark, Paolo Burelli IT University of Copenhagen, Guillaume Andrea Desaphy University of Bari, Alberto Antonio Romano University of Bari, Nicole Novielli University of Bari, Fabio Calefato University of Bari Pre-print | ||
12:22 17mTalk | Developer Prompts in Practice: An Empirical Study of Bias, Security, and Optimization ESEM - Technical Track Dhia Elhaq Rzig University of Michigan - Dearborn, Dhruba Jyoti Paul University of Wisconsin-Madison, Kaiser Pister Univeristy of Wisconsin-Madison, Jordan Henkel Sema4.ai, Foyzul Hassan University of Michigan at Dearborn |