Automating heuristic evaluation of website interfaces using convolutional neural networks

Date of Publication

6-10-2019

Document Type

Master's Thesis

Degree Name

Master of Science in Computer Science

Subject Categories

Computer Sciences

College

College of Computer Studies

Department/Unit

Computer Science

Thesis Adviser

Jordan Aiko P. Deja

Abstract/Summary

Heuristic evaluation is an important phase in both quality assurance and UX de- sign. This ensures that a user interface that is being tested adheres to usability
standards and can be used by any user with ease. Heuristic evaluation typically
takes a long time since it involves consolidating opinions of multiple design ex- perts. This study attempts to automate the detection of usability issues in a given
user interface design to lessen the expense and thereby time needed to hire pro- fessionals and to focus on development-review-revision cycles. The method used
was a data-driven approach through the usage of convolutional neural networks (CNN). A computational model using CNNs to determine whether an interface is good or bad is made from a dataset of screenshots of user interfaces, with a higher
accuracy of that of a simple multilayer perceptron. By comparing the model out- puts to evaluator annotations, several insights regarding the design of e-commerce
websites were also gathered, specifically which heuristics are more important to
the models as compared with which are not as important. The highest perform- ing model yielded 70% accuracy. Further research can lead to fine tuning with a
larger dataset.

Abstract Format

html

Language

English

Format

Electronic

Accession Number

CDTG008215

Keywords

Deep learning (Machine learning); Heuristic programming; Neural networks (Computer science)

Upload Full Text

wf_no

Embargo Period

3-5-2025

This document is currently not available here.

Share

COinS