A COMPUTATIONAL APPROACH TO ELICITING AND MODELING STORIES WITH SOCIAL INTERACTIONS by

In this work, we explored a middle point between the common approach of sharing knowledge using oral narratives and the automatic approach of generation of narratives by computational tools. We describe a new model to capture the flow of events of a story and the changes in the social interactions between the characters as the story progresses. We present a description of the software implementation of the model. Also, we provide a discussion about the design and implementation of the computational tool used to capture stories with the proposed model. The main motivation of this work lies on the fact that each domain of knowledge has people with a wide range of experience and expertise in that domain. Also, within each domain people use narratives as an effective medium of transferring their knowledge. A crucial point is that individuals of a domain have the potential of making valuable contributions to the body of knowledge and that those contributions are often driven in a narrative fashion. In addition, as interactive entertainment continues its role as a pervasive element of today’s culture, the potential of a meaningful experience through the use of narratives can only be achieved if there are tools that aid a wider audience in the creation of those narratives. The AI community has recognized the overwhelming task of authoring stories for interactive entertainment, which demands expertise in computational models for the structure of the story and its execution, as well as expertise in creating the content of the story. Thus, establishing a computational representation of narratives that aid individuals to share their experience and expertise could open the door for capturing the underlying knowledge of a domain.