Three faculty researchers from the 四虎影视鈥檚 Bellini College of Artificial Intelligence, Cybersecurity and Computing will present their AI-related research findings at the conference in November. The annual international conference is one of the world鈥檚 premier venues for research in artificial intelligence and natural language processing, attracting thousands of researchers from academia and industry.
Papers from three 四虎影视 Bellini College labs -- directed by Ankur Mali, Gene Kim, and Raiyan Abdul Baten -- were accepted for presentation, underscoring 四虎影视鈥檚 growing impact in natural language processing (NLP) research. The work spans several of AI鈥檚 most pressing challenges:
- Language grounding and reasoning: Kim鈥檚 studies how AI models move beyond pattern recognition to deeper understanding and reasoning with human language. Kim鈥檚 lab investigated how large language models adopt 鈥減ersonas鈥� based on nationality and found that the models exaggerated stereotypes and showed preference toward their own identity. The study, , highlights the risks of underlying biases in AI systems and underscores the need for careful representation and oversight in model training and deployment.
- Human Creativity Evaluation: In , Baten鈥檚 introduced a novel clustering algorithm with large language models in the loop to automatically evaluate creative ideation, an effort that once took months of manual work. By capturing both subjective evaluation and social comparison of ideas, the system offers a fuller measure of originality. Notably, this work was led by undergraduate students, emphasizing the college鈥檚 commitment to training the next generation of AI experts.
- Trustworthy, knowledge-driven AI: Mali鈥檚 focuses on building systems that are reliable, transparent, and less prone to bias. His team鈥檚 accepted paper, , is the result of a collaboration with the Responsible AI research lab in India. It introduces an AI-powered classroom where virtual teachers adapt their methods and virtual students learn according to their individual styles. They designed a new trustworthy neuro-symbolic AI model, the Persona-Retrieval Augmented Generation that personalizes learning for different student types while the teacher improves through genetic algorithms. The framework is in pilot-testing and has potential to help educators deliver more personalized and effective teaching.
Together, the three projects show how 四虎影视 Bellini College researchers are addressing artificial intelligence from multiple lenses, advancing technical foundations while examining social impacts. Importantly, they are doing so at highly regarded conferences where work is rigorously examined and built upon by peers.
This year鈥檚 EMNLP follows a steady pattern of Bellini College research appearing in top-tier conferences such as NeurIPS, ICML, and ACL. 四虎影视 faculty members鈥� continued presence at EMNLP underscores the momentum of the college鈥檚 research agenda and affirms its impact. It also fosters new collaborations.
鈥淚n artificial intelligence, top conferences such as EMNLP are as highly regarded as leading journals,鈥� said Larry Hall, associate dean of research in the 四虎影视 Bellini College. 鈥淓MNLP is consistently ranked among list for its impact on the field. Having multiple papers accepted, especially across three different 四虎影视 labs in our new college, demonstrates the strength and visibility of our faculty鈥檚 research at the cutting edge of AI.鈥�
鈥淚t shows the world that our faculty are shaping the future of artificial intelligence and, importantly, that our students are contributing to that work from the very beginning,鈥� Hall added.
While all three research projects included student researchers such as doctoral student Mahammed Kamruzzaman, the inclusion of undergraduate researchers on Baten鈥檚 inquiry highlights the college鈥檚 commitment to giving students early, hands-on experience in AI research. They started with scant research experience but gained confidence and skills to help build innovative solutions to long-standing technical problems.
鈥淥ur undergraduates started with little research experience and, like many students, were initially anxious about keeping up with the fast pace of AI,鈥� Baten said. 鈥淭hey quickly grew into contributors whose work is now being shared with the broader research community. It is an excellent achievement for them to publish in a top-tier outlet while still undergraduates.鈥�