Aplicación de técnicas de inteligencia artificial para el desarrollo de interfaces de búsqueda de información
- Garrido Palma, Marcelo
- Antonio Benítez López Doktorvater
Universität der Verteidigung: Universidad Complutense de Madrid
Fecha de defensa: 17 von Oktober von 2013
- Pedro Chacón Fuertes Präsident
- Luis Fernández Moreno Sekretär
- Mari Carmen Marcos Vocal
- Eduardo Bustos Vocal
- Alfredo Burrieza Muñiz Vocal
Art: Dissertation
Zusammenfassung
The research addresses the challenge of multimedia information search. It focuses on the contributions that ontologies bring to the human-computer interaction, specifically to usability improvements achieved by applying semantic metadata in the user interface. This research starts from the fact searching information in the web is relevant phenomenon, and that multimedia search is growing activity among users. Internet reaches more than one third of the world population, where communication between people is one of its main uses. Within this context, search systems have not only improved their ability to crawl, index and retrieve information, but they also have improved their "communication" with users. However, there is still room for improvement. Numerous studies show the difficulties that users have when using search engines. For example, when faced with the task of "querying", or when they have to deal with results that have nothing to do with their needs. Furthermore, authors point out that we are confined to paradigms that have not allowed us to see the search systems of the future; those in which different logics, devices and formats are integrated without losing sight of the users. When it comes to audiovisual, there are additional difficulties besides user interface. These arise from the semantic gap between the raw data and multimedia content with abstract concepts that are used by users to search information. For example, users often have to review full videos in order to find interesting contents, spending most of the time they want. There are also problems when the videos are described in a way that does not makes sense to them, or with words that do not fit with the expected content. In this sense, the task is improving the dialogue between users and systems, providing users with tools that actually facilitate their search, extending the progresses of textual search to the audiovisual world...