Background
As an online advertising marketer, we employ AI technologies to create targeted and efficient advertising campaigns. In Real-Time Bidding (RTB), the demand for automated solutions that dynamically design ad styling and adapt it to the host website is increasing. This aims to maximize reach and audience relevance. The Bachelor thesis will investigate how machine learning methods can be utilized to automate ad styling with CSS and HTML5, enhancing efficiency in RTB environments and enabling flexible adaptation to market changes.
Objective
The goal is to develop a concept and a prototype based on machine learning that automates the styling of online advertisements using CSS and HTML5. One of the objectives is to design ads that appear as native as possible, adapting them to the host website. The focus will be on evaluating and applying various ML methods to enable dynamic adjustments, determining the impact of styling methods on the Click-Through Rate (CTR), and making optimizations where necessary. Additionally, the thesis will analyze the impact of these techniques on reach and campaign efficiency.
Scope of Work
- Analyze the requirements and challenges in RTB as well as existing ML methods for adapting and optimizing ad styles.
- Develop a concept for the automated creation and adjustment of ad styles using machine learning techniques and technologies like CSS and HTML5.
- Crawl and adapt CSS styles and HTML markup from the host website.
- Implement a prototype that tests and evaluates the developed concept in an ad management system.
- Determine and evaluate the influence of ad styles on Click-Through Rate (CTR) and optimize styling methods if needed to improve CTR.
- Conduct tests to measure the effects of automated styling on the reach, relevance, and performance of advertisements.
- Document findings and derive recommendations for applying ML techniques in styling online ads.
Requirements
- Knowledge of machine learning, ideally with experience in applications related to marketing or advertising technology.
- Interest in developing data-driven solutions for the automation and optimization of advertising efforts.
- Basic knowledge of web technologies, especially CSS and HTML5, as well as programmatic advertising is desirable but not essential.
- Analytical skills, creativity, and an independent work approach.
Supervision and Support
- The thesis will be supervised by an experienced team in Ad Tech and AI, providing regular feedback and technical support for developing and implementing the prototype. Application Deadline
- Interested students should submit their application (resume, transcript, and a short motivation letter) by 1.1.2025.
Contact person
Anja Polzer
jobs@definemedia.de
0721 / 6273 999 90