Abstract:
PURPOSE: A data stream distribution parallelization processing service management device and a method thereof are provided to arrange plural tasks effectively by the load property, and to reduce the data input/output load among service tasks by the explosive increase of the input data stream forming the task comprising the service by using a communications means suitable for a task layout structure. CONSTITUTION: A service management unit(412) registers plural tasks comprising a data stream distribution parallelization processing service. A service quality monitoring unit(413) collects load information of the tasks, and load information of plural nodes forming a cluster which provides the data stream distribution parallelization processing service. A scheduling unit(414) distributes and arranges the tasks to the nodes based on the load information. [Reference numerals] (301) Task 1; (302) Task 2; (303) Task 3; (411) Task stream distribution parallelization processing service management device; (412) Service management unit; (413) Service quality monitoring unit; (414) Scheduling unit; (421,431,441) Task operating device; (422,432,442) Task management unit; (423,433,443) Local monitoring unit; (450) Input data stream source; (460) Output data stream source; (AA) Node 1; (BB) Node 2; (CC,DD) Node 3
Abstract:
A system for performing an integrated continuous query process for data stream and stored data by using a user defined shared trigger and a method thereof are provided to reduce time for processing the stored data when integrated continuous queries are processed. A system for performing an integrated continuous query process for data stream and stored data by using a user defined shared trigger comprises a data stream manager(102), an integrated continuous query manager(107), a trigger manager(103), a trigger result manager(104), and an integrated continuous query executer(105). The data stream manager(102) manages the data stream inputted from an external device. The integrated continuous query manager(107) manages the integrated continuous queries inputted from an external application. The trigger manager(103) manages a user defined shared trigger inputted from an external application and registers the shared trigger at an external linked database. The trigger result manager(104) configures and manages trigger result sets resulted from performing the user defined shared trigger registered at the linked database. The integrated continuous query executer(105) processes the integrated continuous queries by referring to the data stream forwarded by the data stream manager(102) and the trigger result sets forwarded by the trigger result manager(104).
Abstract:
PURPOSE: An index and a searching method of high dimensional feature vector data of a Dual Min-LSH algorithm is provided to guarantee QoS to a user by keeping the accuracy over a determined level and reducing the search time. CONSTITUTION: A hash table comprises a Low hash table(226) and a High hash table(228). The Low hash table calculates and stores a MinHash value of higher ordered data on the basis of permutations of arbitrary numbers. The High hash table calculates and stores the MinHash value of higher ordered data on the basis of more permutations including the permutations of the Low hash table for data within a bucket exceeding the number of entries of a threshold in the Low hash table.
Abstract:
1. 청구범위에 기재된 발명이 속한 기술분야 본 발명은 트리거를 이용한 입력 데이터 스트림과 저장 데이터의 통합 질의 처리 시스템 및 그 방법에 관한 것임. 2. 발명이 해결하려고 하는 기술적 과제 본 발명은 연속적으로 입력되는 데이터 스트림과 데이터베이스의 저장 데이터를 통합하여 연속 질의 처리하되, 트리거(trigger)를 이용하여 상기 저장 데이터를 실시간으로 사전 처리해 둠으로써, 통합 질의 처리시 저장 데이터 처리에 소요되는 시간을 줄여 전체 성능이 향상되도록 하기 위한, 트리거를 이용한 입력 데이터 스트림과 저장 데이터의 통합 질의 처리 시스템 및 그 방법을 제공하는데 그 목적이 있음. 3. 발명의 해결방법의 요지 본 발명은, 통합 질의 처리 시스템에 있어서, 외부로부터 연속적으로 입력되는 데이터 스트림을 관리하기 위한 데이터 스트림 관리 수단; 연동 데이터베이스에 트리거를 등록하고, 상기 등록된 트리거의 수행 결과로 결과 집합을 구성하여 실시간 제공하기 위한 트리거 결과 관리 수단; 및 상기 데이터 스트림 관리 수단으로부터 전달받는 데이터 스트림과 상기 연동 데이터베이스의 저장 데이터에 대한 통합 질의를 처리하되, 상기 연동 데이터베이스의 저장 데이터에 대해서는 상기 트리거 결과 관리 수단으로부터 전달받는 결과 집합을 참조하여 통합 질의를 처리하기 위한 실행 수단을 포함한다. 4. 발명의 중요한 용도 본 발명은 데이터 스트림 처리 시스템 등에 이용됨. 통합 연속 질의, XML 데이터 스트림 처리 시스템, 트리거
Abstract:
A system and a method for processing an integrated query for an input data stream and stored data by using a trigger are provided to improve entire performance by reducing time required for processing the stored data when the integrated query is processed, as the data stored in a database is preprocessed in real-time by using the trigger. A data stream manager(102) manages the data stream continuously inputted from the outside. A triggering result set manager(104) registers the triggers(TR1-TRn) to an XML(eXtensible Markup Language) database(103), forms a result set with an execution result of the registered triggers, and provides the result set in real-time. An integrated continuous query executor(105) processes the integrated query by referring to the received result set for the data stored in the XML database while processing the integrated query for the received data stream and the data stored in the XML database. The triggering result manager forms the result set for the triggers automatically executed when the XML database is changed.