A semantic-based toolkit for accurate classification of unstructured data.
Regulation and compliance are pressing and critical concerns in the design and construction industry. Risk management in construction imposes many forms of regulations, including building standards, competing land usage, occupational health and safety issues. The industry is also directly affected by constant environmental changes.
One of the industry’s most complicated segments concerns the construction of large energy and industrial facilities – which require extensive project documentation consisting of thousands of text documents, tables, technical drawings and cost estimates. To improve the quality of project documentation, support and monitoring, an integrated information model is created in a CAD system, which represents all project-related data in a structured manner.
But despite the fact that information models are based on project documentation, they still may contain errors due to manual data entry. These errors may subsequently lead to additional expenses resulting from redundant procurement of materials and components. Additionally, the use of faulty data results in poor equipment designs that man cause integration and installation problems. Errors in data may also develop into health and environment hazards.
In view of the above, it is critical that any discrepancies between project documentation and information models be detected and remedied at the earliest stages of a project. Considering the large amounts of information involved, the best approach to tackle the problem is to use automated intelligent text analysis systems.
ABBYY has developed a technology for intelligent analysis of unstructured textual data that can be used in construction risk management to ensure that data in an information model are identical to the corresponding data in project documentation. The technology is offered as an information extraction module: The ABBYY InfoExtractor SDK.
ABBYY InfoExtractor analyzes text documents and automatically identifies names of objects, specifications and hierarchical relations among system components. Project documents can automatically be converted into structured data to verify the information model. And information extraction results can be visualized and exported into other systems.
Based on ABBYY’s Compreno natural language processing technology, the ABBYY InfoExtractor SDK extracts information with high degrees of precision and recall – enabling a data verification system to detect discrepancies at the early stages of project development, and allowing companies to:
Find out more about ABBYY InfoExtractor: