Unstructured data parsing for structured information
AssigneeNeopost Technologies
InventorKasper Sørensen
Life of this patent
Estimating the term…
Abstract
Systems and methods are provided for computerized, automatic processing of unstructured text to extract bits of contact data to which the extracted text can be linked or attributed. Unstructured text is received and text segments within the text are enriched with metadata labels. A machine learning system is trained on, and used to parse feature values for the text segments and the metadata labels to classify text and generate structured text from the unstructured text.
Claims19
A method comprising:
- receiving unstructured text;
- identifying a plurality of text segments within the unstructured text and enriching the unstructured text by assigning one or more metadata labels to each of the text segments;
- calculating one or more feature values for each segment;
- providing the feature values, metadata labels, and text segments to a trained machine learning system;
- and
- receiving, from the machine learning system, an indication of a classification for at least one of the text segments.
A system comprising:
- a processor configured to:
- receive unstructured text;
- identify a plurality of text segments within the unstructured text and enriching the unstructured text by assigning one or more metadata labels to each of the text segments;
- calculate one or more feature values for each segment;
- provide the feature values, metadata labels, and text segments to a trained machine learning system;
- and
- receive, from the machine learning system, an indication of a classification for at least one of the text segments.
A non-transitory computer-readable storage medium storing a plurality of instructions that, when executed, cause a processor to:
- receive unstructured text;
- identify a plurality of text segments within the unstructured text and enriching the unstructured text by assigning one or more metadata labels to each of the text segments;
- calculate one or more feature values for each segment;
- provide the feature values, metadata labels, and text segments to a trained machine learning system;
- and
- receive, from the machine learning system, an indication of a classification for at least one of the text segments.
Dependencies are inferred from the claim text.