Machine-learning based automated document integration into genealogical trees
AssigneeAncestry.com Operations Inc.
InventorsJack Reese, Luca Lugini, Yingrui Yang, Simon Chu, Gopalkrishna Balkrishna Veni
Life of this patent
Estimating the term…
Abstract
Systems and methods for importing documents are described. An input image is received and preprocessed. OCR and/or page segmentation and chapter detection are performed. Special-case processing is performed for lists, tables, free text, and other categories. Anaphora analysis, stemming, lemmatization, and relationship detection are performed. A genealogical tree is generated, augmented, or merged based on the extracted entities and relationships.
Claims20
A computer-implemented method, comprising:
- detecting, from a digital image, a set of page elements belonging to one or more document structure types of the digital image by utilizing respective document structure detection models corresponding to the one or more document structure types, wherein the respective document structure detection models comprise a form detection model, a table detection model, a list detection model, a genealogical tree detection model, and a free-text detection model;
- segmenting the digital image to generate bounding boxes separating the set of page elements depicted in the digital image;
- extracting a plurality of entities from the set of page elements by implementing an entity extraction model on the bounding boxes;
- determining relationships among the plurality of entities;
- and
- generating genealogy tree data from the plurality of entities and the relationships among the plurality of entities.
A system comprising:
- at least one processor;
- and
- one or more memory devices coupled to the at least one processor, the one or more memory devices storing instructions that, when executed by the at least one processor, cause the at least one processor to:
- detect, from a digital image, a set of page elements belonging go one or more document structure types of the digital image by utilizing respective document structure detection models corresponding to the one or more document structure types, wherein the respective document structure detection models comprise a form detection model, a table detection model, a list detection model, a genealogical tree detection model, and a free-text detection model;
- segmenting the digital image to generate bounding boxes separating the set of page elements depicted in the digital image;
- extract a plurality of entities from the set of page elements by implementing an entity extraction model on the bounding boxes;
- determine relationships among the plurality of entities;
- and
- generate genealogy tree data from the plurality of entities and the relationships among the plurality of entities.
A non-transitory computer readable medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform operations comprising:
- detecting, from a digital image, a set of page elements belonging to one or more document structure types of the digital image by utilizing respective document structure detection models corresponding to the one or more document structure types, wherein the respective document structure detection models comprise a form detection model, a table detection model, a list detection model, a genealogical tree detection model, and a free-text detection model;
- segmenting the digital image to generate bounding boxes separating the set of page elements depicted in the digital image;
- extracting a plurality of entities from the set of page elements by implementing an entity extraction model on the bounding boxes;
- determining relationships among the plurality of entities;
- and
- generating genealogy tree data from the plurality of entities and the relationships among the plurality of entities.
Dependencies are inferred from the claim text.