Browsing by Issue Date, starting with "2014-7-29"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Entropy-Based Dynamic Ad Placement Algorithms for In-Video AdvertisingPublication . Ferreira, Tiago Alexandre Fonseca Ribeiro; Gomes, Abel João PadrãoWith the evolution of the Internet and the increasing number of users over last years, online advertising has become one of the pillars models that sustains many of the Internet businesses. In this dissertation, we review the history of online advertising, will be made, as well as the state-of-the-art of the major scientific contributions in online advertising,in particularly in respect to in-video advertising. In in-video advertising, one of the major issues is to identify the best places for insertion of ads. In the literature, this problem is addressed in different ways. Some methods are designed for a specific genres of video, e.g., football or tennis, while others are independent of genre, trying to identify the meaningful video scenes (a set of continuous and related frames) where ads will be displayed. However, the vast majority of online videos in the Internet are not long enough to identify large scenes. So, in this dissertation we will address a new solution for advertisement insertion in online videos, a solution that can be utilized independently of the duration and genre of the video in question. When developing a solution for in-video advertising, a major challenge rests on the intrusiveness that the ad inserted will take upon the viewer. The intrusiveness is related to the place and timing used by the advertising to be inserted. For these reasons, the algorithm has to take in consideration the "where", "when" and "how" the advertisement should be inserted in the video, so that it is possible to reduce the intrusiveness of the ads to the viewer. In short, in addition to besides being independent of duration and genre, the proposed method for ad placement in video was developed taking in consideration the ad intrusiveness to the user.