Bio：Dr. Vikas Kumar, is a scientific officer in Bioinformatics and Chemoinformatics at the German Federal Institute for Risk Assessment (BfR, Berlin). He also serves as a Senior Scientist at the Health Research Institute (IISPV) and as an Associate Professor at Universitat Rovira i Virgili (URV) in Spain. His work is recognized for its innovative integration of computational models and data science to tackle complex challenges in public health and environmental safety. He is currently co-leading FAIR data and data analytics tasks in the EU PARC project.

Abstract: AI-driven text mining is transforming chemical risk assessment by enabling efficient knowledge extraction and predictive modeling. Within the Partnership for the Assessment of Risks from Chemicals (PARC), AI and Natural Language Processing (NLP) are critical to advancing Adverse Outcome Pathway (AOP) development and Next-Generation Risk Assessment (NGRA). These technologies accelerate the analysis of large-scale scientific and regulatory datasets, enhancing hazard identification, exposure assessment, and chemical safety evaluations. Traditional AOP development relies on keyword-based searches, which often result in fragmented data and time-consuming manual filtering. To address this, PARC's S2CIE (Semantic, Syntactic, and Context-based Information Extraction for AOP development), an advanced information extraction component within AOP-BOT, enables real-time extraction based on semantic, syntactic, and contextual understanding. This significantly reduces query times and improves both relevance and accuracy, allowing researchers to automate literature searches, extract actionable insights, and prioritize evidence for AOP development. PARC plans to integrate a semantically rich rule module for key events, which will facilitate continuous evidence gathering and iterative updates to AOPs as new literature is indexed. Additionally, emerging generative AI models hold great promise for automating AOP hypothesis generation from diverse data sources, further accelerating the development process. This presentation will highlight how S2CIE and AOP-BOT are driving efficiencies in NGRA, improving the transparency, speed, and precision of chemical risk assessment. By integrating AI and NLP into AOP development, these tools represent a significant advancement toward more dynamic, data-driven, and robust chemical safety assurance.