System and method for allocating advertisements
AssigneeIndustrial Technology Research Institute
InventorsTun-Hao You, Yi-Jui Wu, Yi-Jen Yeh
Life of this patent
Estimating the term…
Abstract
According to one exemplary embodiment of an advertisement allocating system, an image object marking and response element receives a TV program signal and then marks at least one object appearing in the TV program during a current time slot; a user element tallies all the users interested in the object, and collects all the feedback information from all the users on the TV program; based on an allowable advertisement set for the current time slot, an advertisement auto-matching element computes the correlation between each specific object and each advertisement of the allowable advertisement set, computes a score for each advertisement of the allowable advertisement set, and extracts an advertisement candidate group for broadcasting.
Claims16
A system for allocating advertisements, comprising:
- an image object marking and response element configured to mark at least one object appearing in a television (TV) program during a current time slot after receiving a TV program signal;
- a user element that executes on a processor to tally all users interested in at least one specific object of said at least one object during said current time slot, and collect all feedback information from said all users for said TV program during said current time slot;
- and
- an advertisement auto-matching element, based on an advertisement set allowed to be broadcasted during the current time slot, that executes on a processor to compute correlation between each of the at least one specific object and each advertisement in said advertisement set, compute a score for each advertisement of said advertisement set, and extract an advertisement candidate group for broadcasting;
- wherein said feedback information includes a combination of one or more than one information of feedback frequency analysis, user interest analysis, and viewer comment of said TV program;
- and
- wherein the feedback frequency analysis includes calculating an operation score of the at least one specific object based on an operation behavior of user feedback, calculating a correlation score between the at least one specific object and at least one advertisement of said advertisement set by comparing a relationship between the at least one object and a description attribute of the at least one advertisement and assigning a numerical value within a normalized scale to the relationship, and computing a score of the at least one advertisement based on the operation score and the correlation score.
A method for allocating advertisements, comprising:
- collecting feedback information of one or more users for a television (TV) program at a first time slot;
- analyzing said feedback information to construct a user feedback set, comparing an advertisement set in accordance with said user feedback set;
- and
- extracting an advertisement candidate group for broadcasting during a second time slot according to said user feedback set and an interest score of each advertisement in said advertisement set;
- wherein said feedback information includes a combination of one or more than one information of feedback frequency analysis, user interest analysis, and viewer comment of said TV program;
- and
- wherein said feedback frequency analysis includes:
- calculating an operation score of each object of at least one object appearing in said TV program according to operation behavior for said TV program in said user feedback set;
- and
- comparing a relationship between said each object of at least one object with at least one advertisement description attribute of each advertisement in said advertisement set, and calculating a correlation score of said each object of at least one object and said each advertisement in said advertisement set by comparing a relationship between said each object of at least one object and said at least one advertisement description attribute of said each advertisement in said advertisement set and assigning a numerical value within a normalized scale to the relationship, and computing the interest score of each advertisement in said advertisement set based on the operation score and the correlation score.
Dependencies are inferred from the claim text.