Social Tagging: Exploring the Image, the Tags, and the Game

An increasing amount of images are being uploaded, shared, and retrieved on the Web. These large image collections need to be properly stored, organized and easily retrieved. Tags have a key role in image retrieval but it is difficult for those who upload the images to also undertake the quality tag assignment for potential future retrieval by others. Relying on professional keyword assignment is not a practical option for large image collections due to resource constraints. Although a number of content-based image retrieval systems have been launched, they have not demonstrated sufficient utility on large-scale image sources on the web, and are usually used as a supplement to existing text-based image retrieval systems. An alternative to professional image indexing can be social tagging -- with two major types being photo-sharing networks and image labeling games. Here we analyze these applications to evaluate their usefulness from the semantic point of view. We also investigate whether social tagging behaviour can be managed. The findings of the study have shown that social tagging can generate a sizeable number of tags that can be classified as interpretive for an image, and that tagging behaviour has a manageable and adjustable nature depending on tagging guidelines