Title:
滞在場所属性特定装置、滞在場所属性特定システム、滞在場所属性特定方法およびプログラム
Document Type and Number:
Japanese Patent JP6405204
Kind Code:
B2
Abstract:
PROBLEM TO BE SOLVED: To determine potent activity series for each staying location attribute, calculate maximum recognition error rate for the determined activity series by using a confusion matrix determined from an activity recognition model, and identify a staying location attribute.SOLUTION: A staying location attribute identification apparatus for identifying a staying location attribute comprises: an activity recognition unit for recognizing a user's activity; an activity history storage unit for storing an activity history by using the recognized activity; an activity pattern storage unit for storing an activity pattern for showing a person's activity characteristic for each staying location attribute; an activity recognition confusion matrix storage unit for storing a confusion matrix corresponding to an activity recognition model; a staying location attribute history storage unit for storing a staying date and time and staying location attribute in association with each other; and a staying location attribute identification unit for identifying a staying location attribute from the activity history by using the activity pattern and the confusion matrix corresponding to the activity recognition model, storing the identified staying location attribute in the staying location attribute history storage unit, and correcting an error of activity series in the activity history.SELECTED DRAWING: Figure 1
More Like This:
Inventors:
Atsushi Matsuno
Keiichiro Hoashi
Keiichiro Hoashi
Application Number:
JP2014241017A
Publication Date:
October 17, 2018
Filing Date:
November 28, 2014
Export Citation:
Assignee:
KDDI CORPORATION
International Classes:
H04M1/00; H04M11/00
Domestic Patent References:
| JP2011040063A | ||||
| JP2010178200A | ||||
| JP2012164227A | ||||
| JP2012003494A | ||||
| JP2014160420A |
Foreign References:
| US20130102283 |
Other References:
Lin Liao et al.,Location-based activity recognition,2006年,URL,http://papers.nips.cc/paper/2911-location-based-activity-recognition.pdf
Attorney, Agent or Firm:
Kiyoshi Kato