Automatic construction of player categories using data clustering techniques

Date of Publication

2009

Document Type

Bachelor's Thesis

Degree Name

Bachelor of Science in Computer Science

Subject Categories

Computer Sciences

College

College of Computer Studies

Department/Unit

Computer Science

Thesis Adviser

Merlin Suarez

Abstract/Summary

There is still much to be explored in the field of player modeling especially when it comes to classifying different players to their respective player categories. This becomes evident as most of the commercial games available today distinguish player types in a relatively shallow manner. These games usually offer users with only a choice of predetermined number of distinct difficulty levels (e.g. easy, medium, hard) which make game progression pre-set and linear. Moreover, most of these games' approaches for basic player categorization still lack an accurate basis. This research aims to provide an approach for automatically creating player categories that are constructed using data extracted from players. Three (3) different data clustering techniques (k-means, Agglomerative Clustering, Neural Networks) will be tested and analyzed to discover the advantages and disadvantages of each when used in creating the data-driven player categories. The test results will be summarized after resulting player categories from each clustering technique are compared and contrasted.

Abstract Format

html

Language

English

Format

Print

Accession Number

TU19847

Shelf Location

Archives, The Learning Commons, 12F, Henry Sy Sr. Hall

Physical Description

1 v. (various foliations) : illustrations (some colored) ; 28 cm.

Keywords

Cluster analysis--Data processing; Data mining--Quality control.

Embargo Period

12-16-2021

This document is currently not available here.

Share

COinS