Soledad Vedovato and Michael Barron, RPI Graduate Students

Soledad Vedovato and Michael Barron, RPI Graduate Students

Michael Garber-Barron

Title: Adaptive Storytelling: Assessing Engagement and Novelty



Storytelling, when happening face to face, is a highly interactive process. A good storyteller attracts the audience by continuously observing their responses and adjusting the storytelling accordingly.

The goal of this project is to simulate this process in digital storyteller. It is an automatic storytelling system that periodically estimates the user’s preferences and adapts the story by balancing novelty and topic consistency.


The structure of the automatic storytelling system is discussed and overviewed. This is then followed by a discussion of the preliminary user studies evaluating the efficacy of the system, and its sub-components.


Soledad Vedovato:

TitleUsing Natural Language to Improve Advanced Search



While Internet search has reached a certain level of adaptation to natural language, its usefulness has been mostly limited to basic searches, quickly defaulting to keyword search or linking to search engine results whenever it faces semantic, syntactic or pragmatic complexity. This behavior leaves behind many of the richest characteristics of human language.


These shortcomings of Internet search engines in understanding natural language in depth have forced users to adopt common maladaptive search strategies such as relying on keywords, pervasive browsing through results, and opting for performing simpler searches than necessary. In addition, their experiences with advanced search have been limited by obstacles such as forms, filters and domain-specific syntax. 


In order to overcome these limitations and demonstrate the usefulness of deeper natural language search, we present advanced search models for four different commonly-used domains: PubMed,, Pinterest and Google Patent Search.