Android source for Shopping Application. Done for VT Information Retrieval course project.

This application uses crowd-sourcing to provide the user with tweets related to current shopping deals in the local area. The application sends HTTP requests (using AsyncTask) to a Hadoop server that stores a corpus of crawled twitter data indexed for shopping relevancy (e.g. store or product names). The server returns a ranked set based on the search query from the application or category of deal selected. The data is returned in XML format and parsed within the app, using XMLPullParser class. The results can be displayed in a textual format, geographical format (using Google Maps API v2), or as a time-series plot of deal trends over a period of time (using AndroidPlot API). The application is composed of two major activities, one for searching, and one for viewing results.  Each activity is further broken down into three fragments (representing standard Android three-tabbed UI) for textual search/results, geographical search/results, and user experience sharing.

Future improvements to this project include integrating it with Twitter so that the user can directly tweet about a shopping experience, providing additional UI components for more complex search queries, crawling and indexing new tweets in real time, and generating dynamic time-series plots.