Abstract:
Apparatus and method for refining subject activity classification for the recognition of daily activities of a subject, and a system for recognizing daily activities using the same. The refining apparatus improves the correctness of subject activity classification using daily activities of a subject, activation time information of sensors mounted on objects associated with the daily activities of the subject, and the suitability of a continuous activity pattern in relation to the daily activities. This improves the correctness of subject activity classification that becomes basic information in daily activity analysis.
Abstract:
일상행위에 대한 이상 징후를 감지하기 위한 방법 및 시스템에 관한 것으로서, 이상 징후 시스템에서 일상행위를 감지하여 미리 저장된 일상행위 정보를 읽어와 이를 기반으로 일상행위열을 생성하고, 상기 일상행위열에서 미리 설정된 서열정렬 알고리즘을 이용하여 이상 징후를 감지하고, 상기 감지된 이상 징후를 사용자에게 서비스함을 특징으로 하며, 이러한 보살핌 서비스에 필요한 이상 징후를 일상행위 패턴의 변화를 통해서 감지함으로써 향후 노인 복지 시설이나, 독거노인 가정에 설치될 보살핌 서비스에 추가하여 사람의 특히 노인의 생활 패턴상의 이상 징후를 효율적으로 감지할 수 있다. 일상행위(ADL) 정보, 이상 징후 감지, 일상행위열(ADLS), 서열 정렬 알고리즘, 감지 서버.
Abstract:
1. 청구범위에 기재된 발명이 속한 기술분야 본 발명은, 시드 단백질 기반 단백질 상호작용 네트워크의 시각화 방법에 관한 것임. 2. 발명이 해결하려고 하는 기술적 과제 본 발명은, 물리적인 관계도가 높은 노드를 중심으로 다단계에 걸쳐 합병을 수행하고 최종 합병 그래프에 대해 다단계에 걸쳐 확장 및 FDP(Force-Directed Placement)를 수행함으로써, 방대한 단백질 상호작용 네트워크를 균형 상태의 그래프로 표현하고 고속으로 시각화하기 위한, 시드 단백질 기반 단백질 상호작용 네트워크의 시각화 방법을 제공하는데 그 목적이 있음. 3. 발명의 해결방법의 요지 본 발명은, 단백질 상호작용 네트워크를 이루는 각 서브 그래프의 노드 리스트를 추출하여 노드 인접 정도에 따라 정렬하는 노드 리스트 정렬 단계; 상기 정렬한 노드 리스트 중에서 노드 우선순위와 타 노드와의 합병 관계에 따라 시드 단백질(Seed Protein)을 선정하는 시드 단백질 선정 단계; 상기 선정한 시드 단백질을 중심으로 인접 노드들을 합병하여 합병 노드를 생성하는 합병 노드 생성 단계; 및 상기 생성한 합병 노드의 초기 위치를 선정하고 해당 시드 단백질을 중심으로 분할포인트 상에 합병된 노드를 위치시킨 후 시각화하는 시각화 단계를 포함함. 4. 발명의 중요한 용도 본 발명은 단백질 상호작용 네트워크의 시각화 등에 이용됨. 바이오인포매틱스, 단백질 상호작용 네트워크, FDP 알고리즘, 인접 노드, 분할포인트, 시각화
Abstract:
본 발명은 무구속 및 무자각의 수면 질 관리 장치에 관한 것으로서, 사용자의 존재 및 움직임에 따른 압력을 감지하는 다수의 압력센서들을 구비하는 무게 센서부; 사용자의 주변 수면 환경을 감지하는 다중 센서를 구비하는 환경 센서부; 및 상기 무게 센서부 및 상기 환경 센서부에서 감지된 정보들을 수집 및 분석하고, 분석된 정보들을 수면의 질에 따라 분류 및 관리하는 제어부를 포함하는 것을 특징으로 하며, 이로써 사용자의 수면 환경을 변경해 가며 학습을 통해 피측정자에 적합한 최적화된 수면 환경을 제공할 수 있으며, 편안하고 안락한 수면을 통한 피로의 회복 및 웰빙을 제공하여 낮 동안의 생활에 활력과 효율성을 제공할 수 있다.
Abstract:
An apparatus and method for refining subject activity classification to recognize an activity of a daily life, and a system for recognizing activity of daily life using the same are provided to increase an accuracy of a behavior classification of a subject person by amending improper behavior among a instantaneous behavior and continuous behavior pattern of the classified subject person. An input buffer receives and stores a behavior classification value of a classified subject person(S200). A first calibrating module extracts and reads a behavior classification value of the subject person stored in the input buffer as the specified interval(S210). The extracted specified interval of a behavior classification value of the subject person is determined whether instantaneous behavior data is in the interval or not(S220). A calibration target data which is an instantaneously behavior data is amended(S230). If no instantaneous behavior data is found, the appropriateness of a continuous action pattern of a behavior classification value of the specified interval of the inside subject person is determined(S240). If an inappropriate pattern is found, an active time band of a object response sensor related to a behavior data of the subject person is considered, and the behavior classification value of the subject person corresponding to the inappropriate pattern is amended(S250). An amended behavior classification value is outputted(S260).
Abstract:
A sensor gateway apparatus based on a ubiquitous sensor network, a system for providing an application service by using the same and a method for providing user circumstance information based on the ubiquitous sensor network are provided to prevent information loss due to unexpected communication disablement and reduce development costs. A sensor data collecting unit(24) collects sensor data about circumstance information of a user measured with at least one sensor adhered to the user or a thing. Sensor data are transmitted to a data management unit(21) from the sensor data collecting unit. The data management unit extracts sensor data corresponding to a predefined data format, a user profile and requirements according to a query of an application service system. A transceiver(27) receives the query of the application service system. The transceiver transmits the sensor data in the application service system.
Abstract:
An abnormality sign sensing method and a system thereof for composing information about the activity of daily life as ADLS according to a time slot are provided to sense the symptom about life pattern of the old people by adding a care service. A sensing server inputs everyday behavior information from ADLS database(210). By using the inputted everyday behavior information, the sensing server produces ADLS(220). The sensing server produces the average ADLS by using ADLS and current ADLS of past(230). The generated ADLS is arranged according to the algorithm(240). A sensing server determines that the symptom is sensed in the life pattern(260). The symptom is informed the nurse and family of to a display unit.
Abstract:
A method for visualizing a protein interworking network around seed protein is provided to express the huge protein interworking network with a stable graph and visualize the protein interworking network at high speed by performing mergence centering on a highly physically related node over a plurality of stages, and performing expansion and FDP(Force-Directed Placement) to the final mergence graph. A node list of each sub-graph forming an protein interaction network is extracted and is arranged according to a node adjacency level(220). Seed protein is selected among the arranged node list according to node priority and mergence relation with other nodes(230). A merged node is generated by merging the adjacent nodes around the selected seed protein(240). An initial position of the merged node is selected and the merged node is located on a division point around the seed protein(260).
Abstract:
A clustering method of gene expressed profile by using gene ontology(GO) is provided to improve biological meaningfulness and reliability of cluster by using the gene ontology as a cluster seed, and minimize loss of information in employment of gene ontology as a cluster seed. A clustering method of gene expressed profile by using gene ontology comprises the steps of: selecting at least one GO term in the gene ontology tree; inputting a gene expression data set; classifying the gene expression data set according to the GO term into each group; firstly clustering the gene expression data belonging to each of the classified groups based on similarity of the gene expression data; and secondly clustering the gene expression data set by using the first clustering results as seed.
Abstract:
A method for reporting new documents relevant to a user profile based on a protein-protein interaction network is provided to prevent excessive information and obtain new information quickly by enabling a user to make the profile based on the protein-protein interaction network, monitoring the new document in an Internet database, and reporting the new document relevant to the user profile in an email. A user makes a user profile by using a protein-protein interaction network and inputting a keyword, an issuance year, an author, an institute, and a journal name to filter documents. A profile processor generates a query by converting a protein ID expressed in the protein-protein interaction network into a protein name and adding the additional filtering information, requests search to a document database, and transfers the query to a document processor when the new document is found in the document database. The document processor extracts protein-protein interaction, checks whether the protein included in the protein-protein interaction network is relevant to the protein ID, makes a summary by highlighting a document number and a part commenting the protein-protein interaction in the document, and reports the summary to the user by using an email.