Context-Aware Advertising Language Analysis and Modeling in Internet Display Ads

Internet display advertising is a substantially growing industry, where advertising spots are dynamically allocated according the product characteristics and the target audience. This is commonly seen on Facebook or Google. However, accurate audience targeting is still an enormous challenge in practice, and often evokes frustration for advertisers and users. However, with the rapid growth and change in this industry, most advertisers have limited knowledge and skills to design advertising campaigns. For example, how to design engaging advertising texts to achieve higher user click rates and to better target the audience. In this project, I proposed an approach to tackle this issue. By analyzing advertising texts along with their contextual information, I will develop advertising language models to automatically generate advertising texts with a given context. This project will enable the partner organization to add a completely new product in this industry and take the advantage in the growing competition.

Faculty Supervisor:

Yvonne Coady

Student:

Dandan Huang

Partner:

Digitoria Media Limited

Discipline:

Computer science

Sector:

Information and communications technologies

University:

University of Victoria