Optimizing personalized web advertising with machine learning

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

School of Business | Bachelor's thesis

Date

2023

Major/Subject

Mcode

Degree programme

Tieto- ja palvelujohtaminen

Language

en

Pages

22+3

Series

Abstract

As time spent on the internet increases, web advertising plays an increasingly important role. The personalization of web advertisements has achieved good results and advertisers are aware of the advantage of a personalized advertisement banner compared to a generic one. This personalization process requires a huge amount of data about the user's online behavior. The desired data contains information about, for example, the user's interests and purchase history, and it is collected from web servers. Machine learning is a great approach for large amounts of raw data and its core methods, supervised, unsupervised and reinforcement learning, can all be used to get the most out of personalized advertisements. This literature review studies the use of these methods in the different stages of the web advertisement personalization process. As a result, it was found that combining different methods produces the best possible result. Another key finding was that a supervised learning method, deep learning, outperforms traditional machine learning techniques in accuracy. The biggest challenges for optimizing personalization are overpersonalization and privacy concerns.

Description

Thesis advisor

Bragge, Johanna

Keywords

machine learning, personalization, web advertising, recommendation

Other note

Citation