Automated classification and localization of daily deal content from the Web

- The identification of effective and computationally inexpensive set of features that are able to discriminate between Web pages that consist of daily deal information and those that do not.
- The construction and systematic comparison of different learning machines based on the identified features that can efficiently classify pages as containing daily deal information with emphasis on both recall and precision.
- The development of a segmentation algorithm capable of determining the number of daily deals present on a given Web page, and accurately localizing those segments of the page that consist of information about one particular deal.