Strategic retrieval of tutorial stories

This paper describes SPIEL, a system for retrieving and presenting tutorial stories for students who are using a social simulation to learn social skills. SPIEL’s task is primarily retrieval, but it requires techniques from casebased reasoning to perform it. SPIEL’s stories are stored in video form, which prevents the use of text-based processing or indexing. Instead of using a story’s text, SPIEL uses complex structured indices intended to represent what the story is about. SPIEL’s tutoring task also differs from strict retrieval because it is not responding to a user’s request for information. The system monitors the student’s actions in a simulated world and brings up relevant stories in the course of that interaction. The standard notion of query does not apply. In its place, SPIEL has a system of storytelling strategies that define what kinds of stories are appropriate at what times. Each strategy makes its own comparison between story and situation, determinations that include but are not limited to assessments of similarity. * This work is supported in part by the Defense Advanced Research Projects Agency, monitored by the Air Force Office of Scientific Research under contract F49620-88-C-0058 and the Office of Naval Research under contract N00014-90-J-4117, by ONR under contract N00014-J-1987, and by the AFOSR under contract AFOSR-89-0493. The Institute for the Learning Sciences was established in 1989 with the support of Andersen Consulting, part of The Arthur Andersen Worldwide Organization. The Institute receives additional support from Ameritech, an Institute partner, and from IBM. Introduction One of the important ways an expert can help a novice is by telling stories and anecdotes (Schank, 1990; Orr, 1986). A computer system for storing and retrieving such stories for pedagogical uses is both a useful educational tool and an interesting challenge for both information retrieval (IR) and casebased reasoning (CBR) research. Finding appropriate material to present to a user is a classic IR problem, but the IR model depends on user initiative: the user must recognize a knowledge need and seek to address it. A tutor, by contrast, must act as well as react: frequently, the student will not know that help is needed. In this paper, I will describe SPIEL (Story Producer for InteractivE Learning), which is testbed for exploring retrieval and presentation strategies for tutorial stories stored on video. SPIEL differs from information retrieval and case retrieval systems in two important ways: ̄ SPIEL combines similarity, dissimilarity, and other types of comparisons to dreate strategic measures of a story’s fitness for retrieval. ̄ Since SPIEL’s stories are in video form, it cannot use text-based indexing methods standard in IR. It uses structured indices more complex than most CBR systems, because it is not storing cases that are themselves intelligible to the system. SPIEL is designed to assist students who are learning social skills. It is embedded in an intelligent learning-by-doing architecture called Guided Social Simulation or GuSS. GuSS provides a social simulation in which 18 From: AAAI Technical Report SS-93-07. Compilation copyright © 1993, AAAI (www.aaai.org). All rights reserved.