Integration of Unstructured Text Documents in A/E/C Model Based Systems

摘要-基于A/E/C模型的系统中非结构化文本文档的集成体系结构、工程和建筑（A/E/C）行业正在采用已建立的通信和信息技术来改善参与建筑项目生命周期各个阶段的许多组织之间的协作、协调和信息交换。目前正在为此目的使用组织间信息系统。它们可以被描述为一组相互关联的组件，这些组件收集、检索、处理、存储和分发信息，以支持不同组织之间的计划、控制和决策制定。在分布式和动态的建筑环境中，交换和集成来自不同来源和不同数据格式的信息的能力对于改进这些系统所支持的流程至关重要。先前的研究表明，在此类系统中集成a /E/C信息的主要障碍与其各种数据类型有关。由于很大比例的施工信息以非结构化文本文档的形式存储，因此管理这类数据的方法对于施工信息管理变得非常重要。本研究的重点是开发方法来集成非结构化文本文档，用于组织间过程，由大型动态分布式虚拟组织在建设项目价值系统中开发。这项调查基于为建筑行业开发的基于互联网的面向项目的组织间信息系统的可用数据。本研究探讨了基于A/E/C模型的系统中非结构化文本文档的自动检索、分类和集成过程。具体来说，分析了信息检索和文本挖掘领域的技术组合，以开发智能搜索引擎来识别与项目模型的每个组件相关的文档。之后，相关文件和项目对象之间的关联将根据IFC规范实现。

Abstract - Integration of Unstructured Text Documents in A/E/C Model Based SystemsThe Architecture, Engineering, and Construction (A/E/C) industry is adopting established communications and information technologies to improve collaboration, coordination, and information exchange among the many organizations that participate in all phases of a construction project life cycle. Inter-organizational information systems, are currently used for this purpose. They can be described as a set of interrelated components that collect, retrieve, process, store, and distribute information to support planning, control and decision making between different organizations. In distributed and dynamic construction environments, the ability to exchange and integrate information from different sources and in different data formats becomes crucial to improving the processes supported by these systems. Previous research showed that a major hurdle in integrating A/E/C information in such systems is related to its variety of data types. Since a large percentage of construction information is stored as unstructured text documents, methods for managing this type of data become important for construction information management.This research focuses on developing methods to integrate unstructured text documents for use in inter-organizational processes, developed by large dynamic distributed virtual organizations in construction project value systems. This investigation is based on available data from internet-based project-oriented inter-organizational information systems developed for the construction industry. In this research, automated processes for retrieval, classification, and integration of unstructured text documents in A/E/C model based systems are explored. Specifically, a combination of techniques from the areas of information retrieval and text mining are analyzed to develop intelligent search engines to identify documents relevant to each component of the project model. Afterwards, an association between relevant documents and project objects will be implemented according to the IFC specification.