ATLAS: Advanced Tourism Planning System

A generic retrieval in Google ignores visitor preferences, time‐updated information about the events taking place in a venue during the visit, and geo-context. Such missing information can easily be found in social networking websites (e.g., Facebook, twitter) and multimedia content (image, video, music) sharing websites (e.g., Flickr, Youtube), where tags disclose users’ behaviour, frequently geo-context, and ratings matched to visitor ones. Geographical information, such as geo‐tag, map, Google street view data, GPS information, user location reveals the relative position to a specific object. Utilizing such information in multimedia semantic analysis enhances location‐based multimedia services, object detection/recognition, and media description/presentation with cultural/region difference. Exploitation of visitor preferences and geographical information will leverage cross‐media/multimedia information retrieval and recommendation for e‐tourism. This is the first advanced functionality ATLAS offers. The underlying concept behind advanced information retrieval/recommendation is to treat it as an optimization problem augmented by meaningful constraints, which will increase precision and thus tourist satisfaction by at least one order of magnitude than current information systems in the tourism domain. Integrating geographical information, online traffic data, and word spotting in streams (e.g., online radio broadcasts, RSS feeds, or twitter) will enable route scheduling to the visitor driver when touring onsite, the second major functionality envisaged within ATLAS. In this project, novel methods and software tools will be developed for: a) geo‐aware social media tagging and annotation, b) geo‐aware multimedia information retrieval and recommendation, c) location‐based services.