System and method for enhancing exploration of data items
OwnerTal Rubenczykinventor
Life of this patent
Estimating the term…
Abstract
The subject matter discloses a method for exploring the most relevant data items, or pieces of information from data sources of various forms and sizes, using the accumulated formalized and clustered behavior of previous users combined with an effective analysis and learning system. According to some embodiments, users are presented with data items that may interest them and are encouraged to rank these data items in order to improve the system's suggestions. The system may also pose the user a selection of questions, meant to further discern the user's preferences.
Claims13
A method;
- the method comprises:
- receiving a plurality of ranks of a plurality of data items;
- wherein said plurality of ranks being ranked by a plurality of users
- calculating an at least one cluster;
- wherein said at least one cluster comprises ranks of a subset of users from said plurality of users and a subset of data items from said plurality of data items;
- wherein data items of said subset of data items being ranked by at least one user from said subset of users;
- wherein said calculating determining said subset of users and said subset of data items in accordance with level of similarity of preference of said subset of data items by said subset of users;
- and
- causing the presenting of an at least one data item from said subset of data items on a computer device of an at least one user from said subset of users.
The method of claim 2, further comprising associating new preferences of said at least one user to said second cluster as a result of said associating a second cluster to said at least one user.
A method for enhancing exploration of data items;
- The method comprises the steps of:
- receiving a plurality of ranks of data items from a user;
- determining a closest cluster to said user;
- said determining being calculated from said plurality of ranks;
- wherein said cluster comprises a subset of said data items;
- and
- presenting to said users highly ranked data items from said cluster.
Dependencies are inferred from the claim text.