Promotion code validation apparatus and method
AssigneeRetailMeNot, Inc.
InventorsDavid Lemphers, Christopher Couhault, Alexandros Suvacioglu
Life of this patent
Estimating the term…
Abstract
A method including interacting with a promotion code provider programmatically, providing a promotion code to a validation function of the promotion code provider, and evaluating a response from the promotion code provider. The act of interacting with a promotion code provider can include controlling an interface that simulates user actions, such as but not limited to controlling an in-memory web browser through an API, where the API provides a set of functions for simulating user actions. Accordingly, the method can be used to quickly and easily determine whether one or more promotion codes, such as online coupons, are valid.
Claims21
1-19. (canceled)
A tangible, non-transitory, machine-readable medium storing instructions that when executed effectuate a machine learning process or processes to form site-level graphs of respective websites suitable to infer paths to respective goal webpages through the respective websites, the operations comprising:
- for each of a plurality of websites, with one or more processors, programmatically mapping paths through the websites by scoring steps along the paths based on likelihoods of the steps along the paths leading to a goal webpage of the respective website, wherein mapping comprises, for each respective website:
- for a plurality of webpages of the respective website:
- requesting, via a network, a webpage of the respective website, the requested webpage being a root webpage of the respective website in an initial iteration of a mapping of the respective website and other webpages in a domain of the respective website, the other webpages being specified by exit elements of webpages in subsequent iterations for the respective website;
- receiving, via the network, the requested webpage;
- determining whether the received webpage is a goal webpage of the respective web site, the goal webpage being a type of webpage on each of the websites different from other types of webpages on each of the websites;
- detecting exit elements in the received webpage, the exit elements specifying other webpages different from the received webpage;
- and
- determining which detected exit elements navigate away from the domain of the respective website;
- forming a site-level graph of the respective website, the site-level graph having a plurality of nodes corresponding to respective webpages of the respective website and a plurality of edges associated with respective pairs of nodes among the plurality of nodes, the edges corresponding to respective exit elements by which the respective website is navigated between webpages corresponding to the respective pair of nodes, wherein forming the site-level graph comprises:
- scoring steps along paths through the respective site-level graph based on a likelihood of the respective step being on a path to a goal node of the respective site-level graph corresponding to the goal webpage of the respective website;
- and
- storing the site-level graph in memory.
A method, comprising:
- for each of a plurality of websites, with one or more processors, programmatically mapping paths through the websites by scoring steps along the paths based on likelihoods of the steps along the paths leading to a goal webpage of the respective web site, wherein mapping comprises, for each respective website:
- for a plurality of webpages of the respective website:
- requesting, via a network, a webpage of the respective website, the requested webpage being a root webpage of the respective website in an initial iteration of a mapping of the respective website and other webpages in a domain of the respective website, the other webpages being specified by exit elements of webpages in subsequent iterations for the respective website;
- receiving, via the network, the requested webpage;
- determining whether the received webpage is a goal webpage of the respective web site, the goal webpage being a type of webpage on each of the websites different from other types of webpages on each of the websites;
- detecting exit elements in the received webpage, the exit elements specifying other webpages different from the received webpage;
- and
- determining which detected exit elements navigate away from the domain of the respective website;
- forming a site-level graph of the respective website, the site-level graph having a plurality of nodes corresponding to respective webpages of the respective website and a plurality of edges associated with respective pairs of nodes among the plurality of nodes, the edges corresponding to respective exit elements by which the respective web site is navigated between webpages corresponding to the respective pair of nodes, wherein forming the site-level graph comprises:
- scoring steps along paths through the respective site-level graph based on a likelihood of the respective step being on a path to a goal node of the respective site-level graph corresponding to the goal webpage of the respective website;
- and
- storing the site-level graph in memory.
Dependencies are inferred from the claim text.