Abstract

Since the inception of the World Wide Web, the amount of data present on websites and internet infrastructure has grown exponentially that researchers continuously develop new and more efficient ways of sorting and presenting information to end-users. Particular websites, such as e-commerce websites, filter data with the help of recommender systems. Over the years, methods have been developed to improve recommender accuracy, yet developers face a problem when new items or users enter the system. With little to no information on user or item preferences, recommender systems struggle generating accurate predictions. This is the cold-start problem. Ackoff defines information as data structured around answers to the question words: what, where, when, who and how many. This paper explores how Ackoff’s definition of information might improve accuracy and alleviate cold-start conditions when applied to the neighborhood model of collaborative filtering (Ackoff, 1989, p. 3).

Library of Congress Subject Headings

Recommender systems (Information filtering)

Publication Date

5-11-2017

Document Type

Thesis

Student Type

Graduate

Degree Name

Information Sciences and Technologies (MS)

Department, Program, or Center

Information Sciences and Technologies (GCCIS)

Advisor

Steve Zilora

Advisor/Committee Member

Qi Yu

Advisor/Committee Member

Jai Kang

Comments

Physical copy available from RIT's Wallace Library at QA76.9.I58 M38 2017

Campus

RIT – Main Campus

Plan Codes

INFOST-MS

Download

Share

COinS