A family of computer systems for delivering individualized advice

作者:

Highlights:

摘要

We describe a prototype framework for a family of individualized-advice systems (IAS). The purpose of an individualized-advice system is to select the most relevant material from a large set of potentially useful documents and present the documents selected to someone seeking advice.IAS systems should not be viewed as applications of artificial intelligence, language recognition, or machine learning. Instead, they are a way to organize a large knowledge base so that the relevant documents can be retrieved. An IAS allows Subject Matter Specialists (SMSs) to classify documents and uses the classification information in the retrieval. The effectiveness of an IAS depends on the ability of the SMS to classify the documents accurately.We illustrate the use of the prototype framework with a detailed description of an application in the area of early childhood development, a prototype system, which we call “Interactive Support System for Early Learning and Inclusion” (ISSELI). ISSELI can offer advice to people who take care of young children who may have developmental issues. This advice suggests safe things that can be done to help the child while the caregivers are waiting for the advice of an expert.A distinguishing property of our approach is that advice-giving systems can be created by the SMSs themselves. The SMSs assemble information and organize it using tables of descriptive text. These tables link the documents to the characteristics of situations where they can be helpful. Neither the creation of the original system nor subsequent in-service maintenance, requires the intervention of someone with programming skills.This paper explains the framework, showing how it can be used by SMSs, and how the resulting systems can be used by the end-user. Examples, taken from ISSELI, illustrate the use of both the framework and the end-user system.

论文关键词:Interactive advice systems,Rough sets,Early childhood education,Special education,Knowledge base

论文评审过程:Received 22 May 2009, Revised 14 February 2010, Accepted 25 February 2010, Available online 10 March 2010.

论文官网地址:https://doi.org/10.1016/j.knosys.2010.02.007