Abstract: A computing server may continuously update a set of nodes that are addable to a data tree based on past interactions of the user with one or more nodes. The computing server may track a recently interacted set of interacted nodes with which the user has interacted within a number of past interactions. The computing server may select a pool of candidate nodes based on the recently interacted set. At least one of the candidate nodes is within a domain boundary of one of the interacted nodes that is in the recently interacted set. The domain boundary may be determined by the degree of relationship. The computing server may present one or more candidate nodes in the pool as a version of the continuously updated set of nodes. The computing server may update the pool of candidate nodes as additional interactions performed by the user updates the recently interacted set.

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.

Abstract: A family tree interface may include a default number of family members in addition to a target node which are expandable upon selection by a user. The default tree interface is expandable by a user vertically to include more generations and laterally. The tree interface includes labels showing a relationship of a tree node to the target node. In some embodiments, one or more family members that have not been rendered may be cached to speed up the visual rendering process. A graphical user interface, in a viewing session, may display an initial view of the family tree associated with the target individual. Upon receipt of an expand request, the viewing session may add the one or more additional family members to generate an expanded view of the family tree. The expanded view may partially adjust the initial view without refreshing the viewing session.

Abstract: Data-sharding systems and/or methods for cost- and time-efficient record search are described. Data-sharding embodiments utilize a name-sharding dimension, optionally in combination with one or more additional dimensions such as record type and year, to reduce latency and reduce search-associated costs. The data-sharding systems and methods embodiments utilize an optimization algorithm to determine a distribution of records related to names. The optimization algorithm may use a three-character prefix for surnames in records to distribute shards across documents, with specific shards relating to no-name and multi-name records allocated.

Abstract: Systems and methods for identifying and segmenting objects from images include a preprocessing module configured to adjust a size of a source image; a region-proposal module configured to propose one or more regions of interest in the size-adjusted source image; and a prediction module configured to predict a classification, bounding box coordinates, and mask. Such systems and methods may utilize end-to-end training of the modules using adversarial loss, facilitating the use of a small training set, and can be configured to process historical documents, such as large images comprising text. The preprocessing module within the systems and methods can utilize a conventional image scaler in tandem with a custom image scaler to provide a resized image suitable for GPU processing, and the region-proposal module can utilize a region-proposal network from a single-stage detection model in tandem with a two-stage detection model paradigm to capture substantially all particles in an image.

Abstract: Disclosed herein relates to example embodiments for recognizing handwritten information in a genealogical record. A computing server may receive a genealogical record. The genealogical record may take the form of an image of a physical form having a structured layout, fields, and handwritten information. The computing server may divide the genealogical record into a plurality of areas based on the structured layout. The computing server may identify, for a particular area, a type of field that is included within the particular area. The computing server may select a handwriting recognition model for identifying the handwritten information in the particular area. The handwriting recognition model may be selected based on the type of the field. The computing server may input an image of the particular area to the handwriting recognition model to generate text of the handwritten information. The computing server may store the text of the handwritten information.

Abstract: A simplified handwriting recognition approach includes a first network comprising convolutional neural network comprising one or more convolutional layers and one or more max-pooling layers. The first network receives an input image of handwriting and outputs an embedding based thereon. A second network comprises a network of cascaded convolutional layers including one or more subnetworks configured to receive an embedding of a handwriting image and output one or more character predictions. The subnetworks are configured to downsample and flatten the embedding to a feature map and then a vector before passing the vector to a dense neural network for character prediction. Certain subnetworks are configured to concatenate an input embedding with an upsampled version of the feature map.

Abstract: Search-result explanation systems, methods, and computer-program products receive a user search query, expand the search query into a plurality of sub-queries, perform a database search using the expanded user search query, and determine which sub-queries of the plurality of sub-queries matched with a particular search result. Results from the database search are re-indexed in an index generated on-the-fly and in-memory, within which the results are searched using the sub-queries to determine matching fields and match types. A score is determined based on the type of match(es) with a particular search result based on one or more predefined weights and normalized using a denominator comprising a fictitious, on-the-fly record configured to receive a perfect score according to the received user search query. A user interface showing ranked results and explanations for the ranking, including a score for the result based on the expanded user search query.

Abstract: Described herein are systems, methods, and other techniques for segmenting an input text. A set of tokens are extracted from the input text. Token representations are computed for the set of tokens. The token representations are provided to a machine learning model that generates a set of label predictions corresponding to the set of tokens. The machine learning model was previously trained to generate label predictions in response to being provided input token representations. Each of the set of label predictions indicates a position of a particular token of the set of tokens with respect to a particular segment. One or more segments within the input text are determined based on the set of label predictions.

Abstract: Systems and methods for determining whether two tree persons in a genealogical database correspond to the same real-life individual. Embodiments include obtaining, from a tree database, a first tree person from a first genealogical tree and a second tree person from a second genealogical tree. Embodiments also include identifying a plurality of familial categories. Embodiments further include, for each familial category of the plurality of familial categories, extracting a first quantity of features for each of the tree persons in the familial category, generating a first similarity score for each possible pairing of tree persons, identifying a representative pairing based on a maximum first similarity score, and extracting a second quantity of features for each of the tree persons in the representative pairing. Embodiments may also include generating a second similarity score based on the second quantity of features.

Abstract: Methods and systems for creating a cluster view person for genealogical studies. Methods may include obtaining a plurality of genealogical trees. Each of the genealogical trees may include a plurality of interconnected nodes representing individuals that are related to each other. Methods may also include identifying one or more of the genealogical trees that contain a similar individual. Whether two individuals are grouped may depend on similarity and/or quality thresholds. Methods may include creating an aggregate individual including each of the similar individuals in each of the identified genealogical trees. The aggregate individual may combine information from each of the similar individuals.

Abstract: Described herein are systems, methods, and other techniques for evaluating a classifier model. The classifier model may be provided with a set of elements to be classified into N classes. Classification results may be obtained from the classifier model. N class-specific precisions and N class-specific recalls for the N classes may be computed based on the classification results. N class-specific precision weights and N class-specific recall weights corresponding to the N classes may be obtained. A weighted f-measure may be computed by weighting the N class-specific precisions with the N class-specific precision weights and weighting the N class-specific recalls with the N class-specific recall weights.

Abstract: Systems and methods for handwriting recognition using language modeling facilitate improved results by using a trained language model to improve results from a handwriting recognition machine learning model. The language model may be a character-based language model trained on a dataset pertinent to field values on which the handwriting recognition model is to be used. A loss prediction module may be trained with the handwriting recognition model and/or the language model and used to determine whether a prediction from the handwriting recognition model should be refined by passing the prediction through the trained language model.

Abstract: Systems, methods, and other techniques for genealogical entity resolution. The systems obtain first tree data and second tree data, the first tree data corresponding to a first tree person and the second tree data corresponding to a second tree person. The systems extract a set of features from the first tree data and the second tree data. The systems further generate an individual-level similarity score for each possible pairing of tree persons based on the set of features to identify a set of most-similar tree persons based on the individual-level similarity score for each possible pairing. The systems also provide a plurality of individual-level similarity scores for the set of most-similar tree persons as input to a family-level ML model to determine that the first tree person and the second tree person correspond to a same individual.

Abstract: Systems and methods for identifying and segmenting objects from images include a preprocessing module configured to adjust a size of a source image; a region-proposal module configured to propose one or more regions of interest in the size-adjusted source image; and a prediction module configured to predict a classification, bounding box coordinates, and mask. Such systems and methods may utilize end-to-end training of the modules using adversarial loss, facilitating the use of a small training set, and can be configured to process historical documents, such as large images comprising text. The preprocessing module within the systems and methods can utilize a conventional image scaler in tandem with a custom image scaler to provide a resized image suitable for GPU processing, and the region-proposal module can utilize a region-proposal network from a single-stage detection model in tandem with a two-stage detection model paradigm to capture substantially all particles in an image.

Abstract: Systems and methods for identifying and segmenting objects from images include a preprocessing module configured to adjust a size of a source image; a region-proposal module configured to propose one or more regions of interest in the size-adjusted source image; and a prediction module configured to predict a classification, bounding box coordinates, and mask. Such systems and methods may utilize end-to-end training of the modules using adversarial loss, facilitating the use of a small training set, and can be configured to process historical documents, such as large images comprising text. The preprocessing module within said systems and methods can utilize a conventional image scaler in tandem with a custom image scaler to provide a resized image suitable for GPU processing, and the region-proposal module can utilize a region-proposal network from a single-stage detection model in tandem with a two-stage detection model paradigm to capture substantially all particles in an image.

Abstract: Described herein are systems, methods, and other techniques for segmenting an input text. A set of tokens are extracted from the input text. Token representations are computed for the set of tokens. The token representations are provided to a machine learning model that generates a set of label predictions corresponding to the set of tokens. The machine learning model was previously trained to generate label predictions in response to being provided input token representations. Each of the set of label predictions indicates a position of a particular token of the set of tokens with respect to a particular segment. One or more segments within the input text are determined based on the set of label predictions.

Abstract: Systems, methods, and other techniques for extracting data from obituaries are provided. In some embodiments, an obituary containing a plurality of words is received. Using a machine learning model, an entity tag from a set of entity tags may be assigned to each of one or more words of the plurality of words. Each particular tag from the set of entity tags may include a relationship component and a category component. The relationship component may indicate a relationship between a particular word and the deceased individual. The category component may indicate a categorization of the particular word to a particular category from a set of categories. The extracted data may be stored in a genealogical database.

Abstract: Techniques for training a machine-learning (ML) model for captioning images are disclosed. A plurality of feature vectors and a plurality of visual attention maps are generated by a visual model of the ML model based on an input image. Each of the plurality of feature vectors correspond to different regions of the input image. A plurality of caption attention maps are generated by an attention model of the ML model based on the plurality of feature vectors. An attention penalty is calculated based on a comparison between the caption attention maps and the visual attention maps. A loss function is calculated based on the attention penalty. One or both of the visual model and the attention model are trained using the loss function.

Abstract: Augmented reality is used to display graphical elements overlaid on a continually updating image of an area around an augmented reality device (e.g., a mobile device). The graphical element may contain geographical location information about a grave of an ancestor and/or biographical information about the ancestor. The continually updating image is captured by a camera of the augmented reality device and updates in response to time and motion of the augmented reality device. Based on orientation data and geographical location data collected by the augmented reality device, the graphical element is updated and displayed on the mobile device.