公开(公告)号：US20160162818A1

公开(公告)日：2016-06-09

申请号：US14861162

申请日：2015-09-22

Applicant: Electronics and Telecommunications Research Institute

Inventor： Tai-Yeon KU ,  Hee-Sun WON

IPC: G06Q10/06

CPC classification number: G06Q10/0633

Abstract: A system for recommending data analysis services is provided. The system comprises: a service catalog storing a list of analysis services; a data catalog storing a list of data of targets for analysis; an algorithm catalog storing a list of analysis algorithms; an interactive analytic tool receiving a user requirement in a word form; an analysis helper receiving the user requirement and extracting an analysis service corresponding to the user requirement from the service catalog; an analysis job generator receiving the user requirement, extracting data and an analysis algorithm corresponding to the user requirement from the data catalog and the algorithm catalog, respectively, and defining the extracted data and the extracted analysis algorithm to one job to generate a job, wherein the analysis job generator transmits the generated job to the analysis helper to deliver the job as recommendation analysis service information to a user through the interactive analytic tool.

Abstract translation: 提供了一种建议数据分析服务的系统。 该系统包括：存储分析服务列表的服务目录; 存储用于分析的目标数据的列表的数据目录; 存储分析算法列表的算法目录; 以单词形式接收用户需求的交互式分析工具; 接收用户要求并从服务目录提取与用户需求相对应的分析服务的分析助手; 接收用户需求的分析作业发生器，分别从数据目录和算法目录提取与用户需求相对应的数据和分析算法，并将所提取的数据和提取的分析算法定义为一个作业以产生作业，其中 分析作业生成器将生成的作业发送到分析帮助器，通过交互式分析工具将作为推荐分析服务信息的作业传递给用户。

公开(公告)号：US20160364211A1

公开(公告)日：2016-12-15

申请号：US14954080

申请日：2015-11-30

Applicant: Electronics and Telecommunications Research Institute

Inventor： Nguyen Minh CHAU ,  Hee-Sun WON

IPC: G06F9/44

CPC classification number: G06F8/20 ,  G06F8/35 ,  G06Q10/06

Abstract: Disclosed are a method of modeling a workflow used in the field of big data and a method and apparatus for executing a workflow model. The method of executing the workflow model according to an embodiment of the present invention comprises receiving a Unified Modeling Language (UML)-based workflow model needing at least one data storage engine and at least one data processing engine, parsing the received workflow model to generate structured information, verifying the validity of the workflow model using the structured information, and transmitting jobs included in the workflow model to data processing engines corresponding to the jobs when it is determined that the workflow model is valid. Thus, developers can be allowed to easily acquire a workflow runnable on various data storage engines and various data processing engines.

Abstract translation: 公开了一种在大数据领域中使用的工作流的建模方法以及用于执行工作流模型的方法和装置。 根据本发明的实施例的执行工作流模型的方法包括：接收基于统一建模语言（UML）的工作流模型，其需要至少一个数据存储引擎和至少一个数据处理引擎，解析接收到的工作流模型以产生 结构化信息，使用结构化信息验证工作流模型的有效性，以及当确定工作流模型有效时，将包括在工作流模型中的作业发送到与作业相对应的数据处理引擎。 因此，开发人员可以容易地获取可在各种数据存储引擎和各种数据处理引擎上运行的工作流程。