Miriam Anschütz, Diego Miguel Lozano, Georg Groh. Proceedings of the 16th International Natural Language Generation Conference. 2023.

Simon Malberg, Roman Poletukhin, Carolin Schuster, Georg Groh Groh. Proceedings of the 5th International Conference on Natural Language Processing for Digital Humanities. 2025.

Edoardo Mosca, Mohamed Hesham Ibrahim Abdalla, Paolo Basso, Margherita Musumeci, Georg Groh. Proceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023). 2023.

As generative NLP can now produce content nearly indistinguishable from human writing, it is becoming difficult to identify genuine research contributions in academic writing and scientific publications. Moreover, information in machine-generated text can be factually wrong or even entirely fabricated. In this work, we introduce a novel benchmark dataset containing human-written and machine-generated scientific papers from SCIgen, GPT-2, GPT-3, ChatGPT, and Galactica, as well as papers co-created by humans and ChatGPT. We also experiment with several types of classifiers—linguistic-based and transformer-based—for detecting the authorship of scientific text. A strong focus is put on generalization capabilities and explainability to highlight the strengths and weaknesses of these detectors. Our work makes an important step towards creating more robust methods for distinguishing between human-written and machine-generated scientific papers, ultimately ensuring the integrity of scientific literature.

Use of novel information and communication technologies are frequently discussed as promising tools to prevent and treat overweight and obesity in children and adolescents. This survey aims to describe the preferences, motives, and needs of children and adolescents regarding nutrition and digital games. We conducted a survey in 6 secondary schools in the southern region of Germany using a 43-item questionnaire. Questions referred to preferences, motives, and needs of children and adolescents regarding nutrition and digital games. In addition, knowledge regarding nutrition was assessed with 4 questions. We collected self-reported sociodemographic and anthropometric data. Descriptive statistical analyses were performed using SPSS. In total, 293 children and adolescents participated in the study, with ages 12-18 years (137 girls, 46.8%), weight 30.0-120.0 (mean 60.2 [SD 13.2]) kg, and height 1.4-2.0 (mean 1.7 [SD 0.1]) m. A total of 5.5% (16/290) correctly answered the 4 questions regarding nutrition knowledge. Study participants acquired digital nutritional information primarily from the internet (166/291, 57.0%) and television (97/291, 33.3%), while school education (161/291, 55.3%) and parents or other adults (209/291, 71.8%) were the most relevant nondigital information sources. Most participants (242/283, 85.5%) reported that they regularly play digital games. More than half (144/236, 61.0%) stated that they play digital games on a daily basis on their smartphones or tablets, and almost 70% (151/282, 66.5%) reported playing digital games for ≤30 minutes without any interruption. One-half of respondents (144/280, 51.4%) also stated that they were interested in receiving information about nutrition while playing digital games. This survey suggests that nutrition knowledge in children and adolescents might be deficient. Most children and adolescents play digital games and express interest in acquiring nutritional information during digital gameplay. A digital game with a focus on sound nutrition could be a potential educational tool for imparting nutrition knowledge and promoting healthier nutrition behaviors in children and adolescents.