-
公开(公告)号:KR1020100045589A
公开(公告)日:2010-05-04
申请号:KR1020080104578
申请日:2008-10-24
Applicant: 한국과학기술원
CPC classification number: G06F17/30734
Abstract: PURPOSE: An inconsistency management method among contextual information in a situation recognition system and a device thereof are provided to keep the reliable context information inside the system by eliminating inconsistency of the contextual information from the situation recognition system. CONSTITUTION: An inconsistency descriptor generator(107) creates a context information inconsistency descriptor. A context information inconsistency manager(108) proceeds detection and removal of inconsistency of the context information. A context information inconsistency detector(110) analyzes the context information inconsistency descriptor. The context information inconsistency detector detects generation of inconsistency about the context information. A context information inconsistency remover(112) eliminates inconsistency of the context information.
Abstract translation: 目的:提供一种情境识别系统及其设备中的上下文信息之间的不一致性管理方法,以通过消除来自情境识别系统的上下文信息的不一致来将系统内的可靠上下文信息保持在系统内。 构成:不一致描述符生成器(107)创建上下文信息不一致描述符。 上下文信息不一致管理器(108)进行检测和去除上下文信息的不一致。 上下文信息不一致检测器(110)分析上下文信息不一致描述符。 上下文信息不一致检测器检测关于上下文信息的不一致性的生成。 上下文信息不一致性去除器(112)消除了上下文信息的不一致性。
-
公开(公告)号:KR1020090039425A
公开(公告)日:2009-04-22
申请号:KR1020070105064
申请日:2007-10-18
CPC classification number: H04L41/042 , H04L41/0233
Abstract: A distributed context aggregation scheme for personal intelligence space using task decomposition scheme is provided to solve a problem of the energy overuse and processing overload by providing a function to distribute the processing task for a coordinator device in the personal intelligence space. A context processing system at a personal intelligent space comprises one coordinator(100) and plural clients(200). The coordinator distributes a sub-task rule to a client device. The coordinator infers upper level context information from a lower level context information which the client device infers according to the sub-task rule. The coordinator comprises a context management framework(110) and a context-aware application(120).
Abstract translation: 提供了一种使用任务分解方案的个人智能空间的分布式上下文聚合方案,通过提供将个人智能空间中的协调器设备的处理任务分配给一个功能来解决能量过度使用和处理过载的问题。 个人智能空间中的上下文处理系统包括一个协调器(100)和多个客户机(200)。 协调器将子任务规则分配给客户端设备。 协调器根据子任务规则从客户端设备推断的较低级别的上下文信息推断上层上下文信息。 协调器包括上下文管理框架(110)和上下文感知应用(120)。
-
公开(公告)号:KR1020090039048A
公开(公告)日:2009-04-22
申请号:KR1020070104457
申请日:2007-10-17
Applicant: 한국과학기술원
IPC: G06F19/28
CPC classification number: G06Q50/10
Abstract: A system for managing context information in the ubiquitous computing environment and a method thereof are provided to avoid a task for processing an unnecessary context event by dispersing the context inference task, concentrated on a global context information management module, to local context information management modules. A plurality of local context information managers(100a-100n) manage the context information about each local domain. The local context information managers recognize and process the context information related to the domain specializing rule of each local domain. A global context information manager(200) manages the context information about a mutual domain, and receives the context information from a corresponding local context information manager to select one of overlapped local context information managers according to the object and kind of the corresponding context information. The global context information manager subscribes the change event of the provided context information.
Abstract translation: 提供了一种用于管理无所不在的计算环境中的上下文信息的系统及其方法,以避免通过将集中于全局上下文信息管理模块的上下文推理任务分散到本地上下文信息管理模块来处理不必要的上下文事件的任务。 多个本地上下文信息管理器(100a-100n)管理关于每个本地域的上下文信息。 本地上下文信息管理器识别和处理与每个本地域的域专业化规则相关的上下文信息。 全局上下文信息管理器(200)管理关于相互域的上下文信息,并且从相应的本地上下文信息管理器接收上下文信息,以根据相应上下文信息的对象和种类来选择一个重叠的本地上下文信息管理器。 全局上下文信息管理器预订所提供的上下文信息的改变事件。
-
公开(公告)号:KR1020070091091A
公开(公告)日:2007-09-07
申请号:KR1020070083373
申请日:2007-08-20
Applicant: 한국과학기술원
CPC classification number: G06F17/3069 , G06F17/15
Abstract: A method for searching semantic resources with semantic space mapping of resource is provided to distribute load among servers by searching the desired resources based on semantic with expandability if many resources are present under large scale ubiquitous computing environment. A first resource description of a landmark representing the reference resources and the second resource description of the desired resources are expressed in ontology with attribute-value relation. Property similarity of attribute/value is calculated by finding the shortest length and a common node height for attribute/value of the first and second resource description, and substituting the shortest length and the common node height for a parameter of a property similarity function. Semantic coordinates of the desired resources are found in a semantic coordinate space by substituting the property similarity of the attribute/value for the parameter of a semantic distance function(S100).
Abstract translation: 提供一种利用资源语义空间映射搜索语义资源的方法,通过在大规模无处不在的计算环境下存在许多资源的情况下,通过搜索具有扩展性的语义来搜索所需资源,从而在服务器之间分配负载。 表示参考资源的标记的第一资源描述和期望资源的第二资源描述在具有属性值关系的本体中表示。 通过找到第一和第二资源描述的属性/值的最短长度和公共节点高度,并将最短长度和公共节点高度替换为属性相似度函数的参数来计算属性/值的属性相似性。 通过将属性/值的属性相似性代入语义距离函数的参数,在语义坐标空间中找到所需资源的语义坐标(S100)。
-
-
-