Skip to content. | Skip to navigation

Personal tools
Sections
Home  /  Research  /  Research Projects  /  Data Representation  /  Semantic Annotation

Semantic Annotation

Overview

Technologies for searching text documents by keyword are abundant and familiar. Some even specialise in searching patent documents. But professional patent researchers must look beyond keywords to find and analyse patents based on a sophisticated understanding of the patent’s content and meaning. Technologies to aid such searches with computer processing are at the cutting edge of information retrieval science.

Matrixware applies semantic analysis tools especially to overcome the difficulties presented by patent documents. Such difficulties include complex technical language, numeric data, contextual relevance based on structure, and sheer volume. These tools are making it possible to search for patents containing, for instance, specific numeric ranges. Or finding patents related to a concept-such as a commercial over-the-counter drug name-that is not explicit in the patent text. In addition, semantic analysis can be an important input for many processing disciplines, such as machine translation, relevance ranking, and data visualisation.

 

Goals

The goals of this project are two-fold: First, to apply semantic text analysis to patent documents in order to aid in discovering information embedded in the meaning of the document, not just the words it contains. Second, to provide a set of open source tools and methodologies that allow searchers to create, hone, organise, store, and use semantic analysis in order to support more accurate document retrieval and analysis.

 

Expected outcome for IP experts

  • A rich set of annotated patent documents to support conceptual and other sophisticated retrieval needs.
  • A rich set of annotated patent documents to support conceptual and other sophisticated retrieval needs.
  • Methodologies for scalable storage and retrieval of annotation data.
  • Customisable interfaces for document search, display, and analysis.

 

Timeline

The project has started in 2007 and is based on annual research cycles.

 

Project Partners

 

Links

Matrixware.net/Semantic Annotation (for more information about methods and findings, as well as publications and related works)

 

Contact

Besides patent search, Semantic Annotation can be applied to various aspects of information management/retrieval. The IRF can provide you with more details about how  Semantic Annotation can help in addressing your concrete needs. Please send your inquiry to: science@ir-facility.org.

MAREC

IRF Scientific Members now have access to the first standardised patent data corpus for research purposes. read more