Information retrieval method and apparatus
First Claim
Patent Images
1. An information retrieval method, comprising the steps of:
- forming a general user model based on statistical data obtained by audience research on the actions of a plurality of users;
said general user model having general user selection taste data;
said statistical data including an audience rating for each of a plurality of genres, a plurality of life-scenes, and wherein each user is classified according to various characteristics including at least an age, a sex, and a life-stage;
inputting a user profile for a specific user;
forming an initial user model for said specific user based on said general user selection data and said user profile;
forming a study user model on the basis of said initial user model and an information selection history for said specific user;
retrieving information suiting said specific user based on said study user model by calculating a genre taste value based on a request time for an electronic program guide (EPG);
said genre taste value being calculated using a life-scene/time function representing a relationship between the life-scene corresponding to the request time and a time-variable coefficient, said time-variable coefficient being determined by a linear interpolation on the basis of the request time and defining a value for said life-scene/time function; and
rearranging and displaying the EPG based on a genre priority table formed using the retrieved information and the calculated genre taste value;
wherein said general user selection taste data is dispersed data, including time related data, that is interpolated into continuous data by an interpolation method specified by an interpolation control identification key.
2 Assignments
0 Petitions
Accused Products
Abstract
Based on a user'"'"'s own basic selection tastes, the priority order of the information is determined, so that an information retrieval method and the apparatus which can retrieve the information desired by the user easily among from a huge amount of programs can be realized.
-
Citations
6 Claims
-
1. An information retrieval method, comprising the steps of:
-
forming a general user model based on statistical data obtained by audience research on the actions of a plurality of users;
said general user model having general user selection taste data;
said statistical data including an audience rating for each of a plurality of genres, a plurality of life-scenes, and wherein each user is classified according to various characteristics including at least an age, a sex, and a life-stage;inputting a user profile for a specific user; forming an initial user model for said specific user based on said general user selection data and said user profile; forming a study user model on the basis of said initial user model and an information selection history for said specific user; retrieving information suiting said specific user based on said study user model by calculating a genre taste value based on a request time for an electronic program guide (EPG);
said genre taste value being calculated using a life-scene/time function representing a relationship between the life-scene corresponding to the request time and a time-variable coefficient, said time-variable coefficient being determined by a linear interpolation on the basis of the request time and defining a value for said life-scene/time function; andrearranging and displaying the EPG based on a genre priority table formed using the retrieved information and the calculated genre taste value; wherein said general user selection taste data is dispersed data, including time related data, that is interpolated into continuous data by an interpolation method specified by an interpolation control identification key. - View Dependent Claims (2)
-
-
3. An information retrieval apparatus, comprising:
-
general user model forming means for forming a general user model based on statistical data obtained by audience research on the actions of a plurality of users;
said general user model having general user selection taste data;
said statistical data including an audience rating for each of a plurality of genres, a plurality of life-scenes, and wherein each user is classified according to various characteristics including at least an age, a sex, and life-scene;input means for inputting a user profile for a specific user; initial user model forming means for forming an initial user model for said specific user based on said general user selection data and said user profile; study user model forming means for forming a study user model on the basis of said initial user model and an information selection history for said specific user; retrieving means for retrieving information suiting said specific user based on said study user model by calculating a genre taste value based on a request time for an electronic program guide (EPG);
said genre taste value being calculated using a life-scene/time function representing a relationship between the life-scene corresponding to the request time and a time-variable coefficient, said time-variable coefficient being determined by a linear interpolation on the basis of the request time and defining a value for said life-scene/time function; anddisplay means for rearranging and displaying the EPG based on a genre priority table formed using the retrieved information and the calculated genre taste value; wherein said general user selection taste data is dispersed data, including time related data, that is interpolated into continuous data by an interpolation method specified by an interpolation control identification key. - View Dependent Claims (4, 5, 6)
-
Specification