Back to results
US 10,762,119 B2Granted
Semantic labeling apparatus and method thereof
AssigneeSamsung Electronics Co., Ltd.
InventorsJae Mo Sung, Min Young Mun
Life of this patent
Estimating the term…
Abstract
A semantic labeling apparatus and method thereof include a place identifier processor configured to, based on location data of a user, generate place attributes of places that indicate information of a user visit for each place, wherein user location remains unchanged within the places for a predetermined period of time. A group identifier processor is configured to cluster the places based on the place attributes, classify the places into groups, acquire a semantic label for each of the groups, and designate the acquired semantic label as the semantic label of each of the groups. A label determiner is configured to determine the semantic label of each of the groups as a semantic label of each member place of each of the groups.
Claims21
A semantic labeling apparatus for estimating a service, comprising:
- a place identifier processor configured to, based on location data of a user, generate place attributes of places that indicate information of a user visit for each place of the places the user visits, wherein user location remains unchanged within the places for a predetermined period of time;
- a group identifier processor configured to cluster the places based on the place attributes, classify the places into groups, select, as a representative place, one member place of each of the groups based on derivative information being generated using the place attributes of the places, the generated derivative information including an accumulated number and time of user visits to the one member place during a day of a week, the one member place being determined to have a largest number of user visits of the places for another predetermined period of time, acquire a semantic label of the selected representative place, and designate the acquired semantic label of the representative place as the semantic label for a corresponding group;
- a label determiner processor configured to determine the designated semantic label for the corresponding group as a semantic label of each member place in the corresponding group, wherein
- in response to a current location of the user being determined to be a new place based on attributes of the new place, the label determiner processor is configured to estimate a semantic label of the new place based on a result of comparing the attributes of the new place with the generated place attributes of the places;
- and
- an estimator processor configured to estimate a service that the user is determined likely to need based on the determined semantic label for the corresponding group associated with the location data of the user and based on the estimated semantic label of the new place for the current location of the user, as determined by the label determiner processor.
3 more elements
A processor-implemented semantic labeling method of estimating a service, comprising:
- generating, based on location data of a user, place attributes of places that indicate information of a user visit for each place of the places the user visits, wherein user location remains unchanged within the places for a predetermined period of time;
- clustering the places based on the place attributes, classifying the places into groups, selecting, as a representative place, one member place of each of the groups based on derivative information being generated using the place attributes of the places, the generated derivative information including an accumulated number and time of user visits to the one member place during a day of a week, the one member place being determined to have a largest number of user visits of the places for another predetermined period of time, acquiring a semantic label of the selected representative place, and designating the acquired semantic label of the representative place as the semantic label for a corresponding group;
- determining the designated semantic label for the corresponding group as a semantic label of each member place in the corresponding group, wherein
- in response to a current location of the user being determined to be a new place based on attributes of the new place, estimating a semantic label of the new place based on a result of comparing the attributes of the new place with the generated place attributes of the places;
- and
- estimating a service that the user is likely to need based on the determined semantic label for the corresponding group associated with the location data of the user and based on the estimated semantic label of the new place for the current location of the user.
3 more elements
A processor-implemented semantic labeling method of estimating a service, comprising:
- generating, based on location data of a user, place attributes of places that indicate information of a user visit for each place of the places the user visits, wherein user location remains unchanged within the places for a predetermined period of time;
- clustering the places based on the place attributes;
- classifying the places into groups;
- selecting, as a representative place, one member place of each of the groups based on derivative information being generated using the place attributes of the places, the generated derivative information including an accumulated number and time of user visits to the one member place during a day of a week, the one member place being determined to have a largest number of user visits of the places for another predetermined period of time;
- acquiring a semantic label of the selected representative place;
- designating the acquired semantic label of the representative place as the semantic label for a corresponding group;
- determining the designated semantic label for the corresponding group as a semantic label of each member place in the corresponding group;
- and
- in response to a current location of the user being determined to be a new place based on attributes of the new place, estimating a semantic label of the new place based on a result of comparing the attributes of the new place with the generated place attributes of the places.
6 more elements
Dependencies are inferred from the claim text.