US 2020/0050662 A1 · System And Method To Generate A Labeled Dataset For Training An Entity Detection System

Abstract

“Semi-supervised” machine learning relies on less human input than a supervised algorithm to train a machine learning algorithm to perform entity recognition (NER). Starting with a known entity value or known pattern value for a specific entity type, phrases in a training data corpus are identified that include the known entity value. Context-value patterns are generated to match selected phrases that include the known entity value. One or more context-value patterns may be validated based on human input. The validated patterns identify additional entity values. A subset of the additional entity values may also be validated based on human input. Occurrences of validated entity values may be labeled in the training corpus. Sample phrases from the labeled training dataset may be extracted to form a reduced-size training set for a supervised machine learning model which may be further used in production to label data for any named entity recognition application.

Claims36

Dependent claims

A non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising:

identifying a plurality of citations of a seed value for an entity type in a training data corpus;
determining a context-value pattern for the plurality of the citations of the seed value for the entity type;
identifying a plurality of candidate values, for the entity type, in the training data corpus based on each of the plurality of candidate values being cited in the training data corpus in accordance with the context-value pattern;
and
labeling at least a subset of the plurality of candidate values with the entity type.

2 more elements

A system comprising:

at least one hardware device including a processor;
and
the system configured to perform operations comprising:
identifying a plurality of citations of a seed value for an entity type in a training data corpus;
determining a context-value pattern for the plurality of the citations of the seed value for the entity type;
identifying a plurality of candidate values, for the entity type, in the training data corpus based on each of the plurality of candidate values being cited in the training data corpus in accordance with the context-value pattern;
and
labeling at least a subset of the plurality of candidate values with the entity type.

5 more elements

Dependencies are inferred from the claim text.

System And Method To Generate A Labeled Dataset For Training An Entity Detection System

Life of this patent

System And Method To Generate A Labeled Dataset For Training An Entity Detection System

Life of this patent

Abstract

Claims36