About me
Shikha Soneji is a doctoral candidate in Informatics at Penn State, where her research lies at the intersection of human-centered AI, privacy, and policy understanding. Her work focuses on building and evaluating AI systems that help people interpret complex, high-stakes documents, such as privacy policies and terms of service. She develops hierarchical ontologies, NLP- and LLM-based classification methods, and mixed-methods evaluation frameworks to examine both technical performance and how real users perceive, trust, and act on AI-generated explanations.
Shikha's research has been published multiple times in venues focused on AI for social impact, where her work has been recognized with awards for its contribution to improving transparency, fairness, and user understanding in digital systems. A central theme of her research is bridging technical innovation with societal needs, particularly in domains where information asymmetry can meaningfully affect user rights and decision-making.
In addition to her research, Shikha is deeply committed to mentorship. She has mentored undergraduate and graduate students, as well as early-career researchers, supporting them in research design, technical skill development, and navigating interdisciplinary academic paths. She is especially passionate about mentoring students from diverse backgrounds and helping them translate abstract ideas into impactful, real-world research.
Upon graduating from Penn State, Shikha aspires to continue working at the intersection of research and practice, advancing responsible, user-centered AI systems that are transparent, trustworthy, and grounded in social impact.
