Unlock instant, AI-driven research and patent intelligence for your innovation.

A hierarchical entity recognition and semantic modeling framework for information extraction

An entity recognition, entity technology, applied in informatics, semantic analysis, healthcare informatics, etc., can solve problems such as difficulty in incorporating clinician knowledge into extraction tasks, and inappropriate application of clinician professional knowledge and understanding technology.

Pending Publication Date: 2020-09-11
KONINKLJIJKE PHILIPS NV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Furthermore, current approaches make incorporating clinician knowledge into the extraction task difficult, and clinician expertise and understanding techniques are not properly applied to machine learning schemes

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A hierarchical entity recognition and semantic modeling framework for information extraction
  • A hierarchical entity recognition and semantic modeling framework for information extraction
  • A hierarchical entity recognition and semantic modeling framework for information extraction

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] Extracting meaningful entities from documents (especially when entities are distributed within the document in a nested fashion) can often prove to be a difficult task to automate. In some instances, a rule-based system can be employed, which is easy to provide but is often overly restrictive in the entities that can be extracted by any particular rule. For example, although a user may identify a series of rules for extracting entities in a particular domain, the user will often not have contingent rules for every possible variation of terms that may describe an entity. Consequently, rule-based approaches are often not to scale. Machine learning methods can be used as an alternative for extracting entities. However, machine learning methods require "ground truth"; that is, known correct answers that can be used for training and later verifying the performance of the machine learning model.

[0022] When taken alone, both rule-based systems and machine learning model e...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses extracting entities from a document with a hierarchical entity graph of entities. Entity definitions and entity recognition definitions are customized by a user and provided. The configuration information is utilized to generate (905) an entity graph, which is then utilized to parse one or more documents. In some implementations, the resulting parse tree may be utilized, inconjunction with user feedback, to generate one or more training instances for a machine learning model assigned to one or more of the custom nodes as an entity recognition definition. Parsing of theresulting tree may be performed with a lazy parsing methodology, with only the portions of interest to the user being identified in the document.

Description

Background technique [0001] Information extraction (particularly in clinical documents) often requires the collection of key-value pairs of interest to clinicians. To accomplish this task, entity recognition is used to identify keys, and secondly, relationships can be extracted from the identified entities in order to identify meaningful entities in the document. Among the conventional methods for extracting meaningful related entities, two approaches are generally employed: namely, entity extraction (ie, NER) and entity-relationship extraction (ie, ER). Existing approaches require a set of manual (eg, user-curated) rules or heuristics, often in the form of regular expressions, to identify entities in documents. Regular expressions are useful in many ways because they are quick to create and require little data to run and test samples, and also because they are concise in generation and presentation. However, regular expressions are still limited in their simplicity when use...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G16H15/00G06K9/72G06F40/14G06F40/295
CPCG16H15/00G06F40/169G06F40/205G06F40/295G06F40/30G06F18/24143G06N20/00G06F40/284G06F40/242G06F40/154G06F40/211G06F40/289
Inventor 胡意仪欧阳恩李作峰
Owner KONINKLJIJKE PHILIPS NV