The proposed project explores the problem of automated narrative generation, the creation of narratives by computer systems. The project introduces a transformative new approach to narrative generation that blends human and computational creativity with crowdsourcing. The system addresses fundamental limitations of computer reliance on pre-coded domain knowledge in order to generate a virtually unlimited variety of narratives and make it possible for non-experts and non-programmers to create interactive narratives. The research has four major components: (1) Develop artificial intelligence algorithms that emulate human ability to create narratives. (2) Design and implement novel models of human-computer creative collaboration. (3) Study fundamental questions pertaining to human narrative learning and cognition. (4) Explore the role of narrative generation in real-world domains: virtual agents that create rapport with humans and intelligent creativity augmentation tools for creating and sharing interactive experiences. The work will be piloted in two healthcare systems: a virtual agent that creates rapport and fosters longitudinal engagement with patients through autobiographical narratives; and intelligent tools that allow caregivers to create social skill scenarios for young adults with autism to practice.
Narratives are important because they are a fundamental means by which humans organize, understand, and explain the world. If computer systems could create effective narratives, they would be better able to interact with people. The research will result in novel algorithms, software, and a body of experimental knowledge that will enable the building of interactive narrative systems that are practical, scalable, usable by non-programmers, and can address societally important problems in education, training, and healthcare interventions. The proposed approach to creativity support will significantly lower the technical, artistic, and skill barriers to creating interactive narrative systems, opening avenues for educators, trainers, caregivers, and hobbyists to create and share interactive experiences.
The project includes an educational plan to develop a sustainable, annual summer hack-a-thon camp wherein high school students work alongside K-12 teachers to create interactive narratives that motivate and guide classroom inquiry based learning. The hack-a-thon aims to provide minority and low socioeconomic high school students with hands-on computing science experience and to produce a library of interactive inquiry-based learning software systems for K-12 teachers.
Interagency Autism Coordinating Committee (IACC)
Autism Research Database (AFD)
