سال انتشار: ۱۳۸۴

محل انتشار: یازدهمین کنفرانس سالانه انجمن کامپیوتر ایران

تعداد صفحات: ۴

نویسنده(ها):

Alireza Vazifedoost – Database Research Group Faculty of Electrical andComputer EngineeringSchool of Engineering,University of Tehran
Ashkan Bayati –
Farhad Oroumchian – Department of Computer Science University of Wollongong in Dubai
Hassan Seyedrazi – Faculty of Electrical and Computer Engineering School of Engineering, University of Tehran

چکیده:

This paper describes the usage of a new theory for combining various evidence provided by Plausible inferences in Information Retrieval (PLIR). PLIR is an experimental information retrieval system based on applying human like inferences, with plausible reasoning, for inferring relevancy of a document to a user’s query. It’s often more than one inference; hence many evidence concluding the relevancy with different degrees of uncertainty. The Current method used is Depmster-Shafer (DS) theory of evidence for combining the evidences. This paper highlights a few problems that rise when using DS, and then as a remedy Dezert- Smarandache’s theory is introduced.