초록 ▼

A method and apparatus are disclosed for recommending television programming using decision trees. Inductive principles are utilized to identify a set of recommended programs that may be of interest to a particular viewer, based on past viewing behavior. A user's viewing history is monitored and the

A method and apparatus are disclosed for recommending television programming using decision trees. Inductive principles are utilized to identify a set of recommended programs that may be of interest to a particular viewer, based on past viewing behavior. A user's viewing history is monitored and the shows that are actually watched by a user (positive examples) and the shows that are not watched by the user (negative examples) are analyzed. For each positive and negative program example (i.e., programs watched and not watched), a number of program attributes are classified in the user profile. The various attributes are then positioned in a hierarchical decision tree based on a ranking of the entropy of each attribute. Each node and sub-node in the decision tree corresponds to a given attribute from the user profile. Each leaf node in the decision tree corresponds to either a positive or negative recommendation for a program mounted at the corresponding leaf node. The decision tree is built using a decision tree process that implements a “top-down divide and conquer” approach. The decision tree can thereafter be applied to an electronic program guide to make program recommendations.

대표청구항 ▼

A method and apparatus are disclosed for recommending television programming using decision trees. Inductive principles are utilized to identify a set of recommended programs that may be of interest to a particular viewer, based on past viewing behavior. A user's viewing history is monitored and the

A method and apparatus are disclosed for recommending television programming using decision trees. Inductive principles are utilized to identify a set of recommended programs that may be of interest to a particular viewer, based on past viewing behavior. A user's viewing history is monitored and the shows that are actually watched by a user (positive examples) and the shows that are not watched by the user (negative examples) are analyzed. For each positive and negative program example (i.e., programs watched and not watched), a number of program attributes are classified in the user profile. The various attributes are then positioned in a hierarchical decision tree based on a ranking of the entropy of each attribute. Each node and sub-node in the decision tree corresponds to a given attribute from the user profile. Each leaf node in the decision tree corresponds to either a positive or negative recommendation for a program mounted at the corresponding leaf node. The decision tree is built using a decision tree process that implements a “top-down divide and conquer” approach. The decision tree can thereafter be applied to an electronic program guide to make program recommendations. inus;X| between said point X and said PiD; determining whether said point Piis acceptable for said inversion of said point X by comparing said magnitude of the difference to a tolerance δ; and, selecting said tolerance δ as a fraction of a width of a pixel for a computer display surface wherein said fraction includes one half;c) locating a next approximation point Pi+1for said inversion of said point X if said approximation point Piis not acceptable for said inversion of said point X; and,d) repeating steps (b) and (c) until said approximation point is acceptable for said inversion of said point X.2. The method of claim 1 wherein said undistorted surface is included in said detail-in-context presentation.3. The method of claim 1 and further comprising the step of constructing a line RVP-X from a point RVP above said undistorted surface, through said point X, and through said undistorted surface to locate said first approximation point Piat a point of intersection of said line RVP-X and said undistorted surface.4. The method of claim 3 wherein said point RVP is a reference viewpoint for said detail-in-context presentation.5. The method of claim 3 further comprising the steps of:a) projecting said point PiDonto said line RVP-X to locate a point PiP, where said point PiPis a closest point to said point PiDon said line RVP-X; and,b) projecting said point PiPonto said undistorted surface in a direction opposite to that of a displacement due to distortion to locate said next approximation point Pi+1for said inversion of said point X, wherein said displacement due to distortion is given by a line Fo-F constructed through said undistorted surface and a focus F of said distorted surface, and wherein said point Pi+1is located on said undistorted surface at a point of intersection of said undistorted surface and a line constructed parallel to said line Fo-F and passing through said point PiP.6. The method of claim 1 and further comprising the step of bisecting said point Pito counter divergence in successive approximations of said point Pidue to folds or discontinuities in said distorted surface.7. The method of claim 1 wherein said distorted surface is a plane.8. The method of claim 1 wherein distorted surface is defined by said distortion function D.9. The method of claim 8 wherein said distortion function D is an n-dimensional function, where n is an integer greater than zero.10. The method of claim 9 wherein said distortion function D is a three-dimensional function.11. The method of claim 8 wherein said distortion function D is a lens function.12. A system for inverting a distorted surface in a detail-in-context presentation, said system having memory, a display, and an input device, said system comprising: a processor coupled to said memory, display, and input device and adapted for:(a) locating a first approximation point Pifor inversion of a point X, wherein said point Pion an undistorted surface and said point X is on said distorted surface;(b) obtaining a point PiDby displacing said point Pionto said distorted surface by applying a distortion function D; calculating a magnitude of the difference |PiD−X| between said point X and said point PiD; determining whether said point Piis acceptable for said inversion of said point X comparing said magnitude of the difference to a tolerance δ; and, selecting said tolerance δ as a fraction of a width of a pixel for said display wherein said fraction includes one half;(c) locating a next approximation point Pi+1for said inversion of said point X if said approximation point Piis not acceptable or said inversion of said point X; and,(d) repeating (c) until said approximation point is acceptable for said inversion of said point X.13. The system of claim 12 wherein said undistorted surface is included in said detail-in-context presentation.14. The system of claim 12 wherein