Development of Display Ads Retrieval System to Match Publisher's Contents
This addresses the problem of optimizing ad relevance for publishers in internet advertising, but appears incremental as it builds on existing methods without claiming major breakthroughs.
The paper tackled the challenge of matching contextual ads to web page content by developing a prototype retrieval system, demonstrating a working mechanism for selecting ads that align with publisher content.
The technological transformation and automation of digital content delivery has revolutionized the media industry. Advertising landscape is gradually shifting its traditional media forms to the emergent of Internet advertising. In this paper, the types of internet advertising to be discussed on are contextual and sponsored search ads. These types of advertising have the central challenge of finding the best match between a given context and a suitable advertisement, through a principled method. Furthermore, there are four main players that exist in the Internet advertising ecosystem: users, advertisers, ad exchange and publishers. Hence, to find ways to counter the central challenge, the paper addresses two objectives: how to successfully make the best contextual ads selections to match to a web page content to ensure that there is a valuable connection between the web page and the contextual ads. All methods, discussions, conclusion and future recommendations are presented as per sections. Hence, in order to prove the working mechanism of matching contextual ads and web pages, web pages together with the ads matching system are developed as a prototype.