WWW/Internet - Portal do Software Público Brasileiro

ISBN: 978-972-8939-25-0 © 2010 IADISto capture Wi-Fi signals and to store the profiles in the mobile devices, we used the OpenNETCF library andthe .NET System.Xml library, respectively.The integration between the TrailM and the UbiTrade prototypes was mainly based on changes made inOpportunity Consultant in order to access the services provided by the Web services of TrailServices. TheTrailM Client prototype was developed through the addition of new functionalities in Personal Assistant.The UbiTrade location system is based on triangulation of Wi-Fi antennas. Personal Assistant obtains theantennas’ powers and forwards to Location System through a Web service. The system uses this informationto infer the mobile device location. The locations must be previously registered on the system database. Themobile device location is inferred through a comparison between the data sent by Personal Assistant and thedata previously captured through samples.4.2 Case Study4.2.1 Obtainment of the TrailsThe case study was based on simulated generation of two dealers’ trails, which were used in the real scenariodescribed in section 4.2.2.Figure 6. Relationship of the real and simulation environment mapsFigure 6 presents the relation between the real and simulation environments used in the trails generation,which is composed of the Houses, the Metro, and the Shopping regions. For example, a movement fromroom 215 to room 206 represents the dealer’s presence on the train, moving from House to Shopping.We defined these regions because we wanted to monitor the dealers’ activities while they were movingfrom their houses to the shopping. Within the regions, we mapped some location where the users could go,for instance, the cabstand, the fast food 1, and station 2.The Metro region represents the area in which the dealers move from their houses to the shopping andvice versa, therefore, the system can record the events made by the dealer during the movement between thestations. The third region corresponds to the shopping. Within the shopping region, we defined twosubregions representing the stores and the food court. A region can be composed of one or more subregions,and each subregion contains at least one location with the values of its geographical position.Table 2. Example of the recordsDealer Device Event Location Extension DateCládio HTC 4351 Entry into location Shopping - Feb. 6 08:03Cládio HTC 4351 Offer inclusion Jewelry Store Title: PianoFeb. 6 12:01Category: Instrument SoundJoão HTC 4351 Entry into location Shopping - Feb. 6 19:32Cládio HTC 4351 Exit from location Jewelry Store - Feb. 6 20:03João HTC 4351 Created Opportunity Shopping DO Notification Feb. 6 21:12In the simulation, the dealers’ profile information was based in two real people, focusing their preferencesand interests related to desires and offers, and registered in UbiTrade. The trails generation was based on the24