[LarKC Project]BOTTARI – Semantic Location-based Social Media Analytics
BOTTARI – Description of the application
BOTTARI is a service which provides geo context information and social context information. The service recommends Points of Interest (POIs) with geo context information for users through augmented reality based on location and view angle of users’ smart phones (Android OS).
Geo context information includes detailed information of the POIs such as GPS coordinates, contacts, menus, price and so on, and also includes subjective information such as ambience, reviews, etc. Social context information includes reputations of POIs analysed from social media messages by machine learning. And BOTTARI is able to recommend some POIs based on the information of social context information. For these functions, there are four buttons (Figure 1). ‘For me’ to recommend POIs which can be preferred by the current user, ‘Popular’ ones which have much of positive reputations on Twitter, “Emerging” ones which is similar to “Popular” showing the most popular in the last 6 months and ‘Interesting’ ones which can be interesting to visit for the foreign visitors in Korea.
From the BOTTARI user interface, POIs are shown on the basis of the current user position. And if the menu button located left bottom in the UI is pressed, it will display some options for choosing recommendation and distance filters (shown in Figure 1).
If one of the POIs is selected, BOTTARI will display its description/introduction and the list of POIs which are near the selected POI, and there is the ‘Details’ button to query details of the POI (see Figure 2). If the button is selected, details are displayed including reputations, map and pictures (shown in Figure 3).
In the details view, the “reputation” button let the user see the details on the POI reputation (Figure 4). The reputation trends view shows monthly reputations from some perspectives of overall, taste, comfort and service using graphs displaying positive and negative messages counts.