Emc IP Holding Company LLC

First American Financial Corporation

Emc Corporation

Jeffrey Norton

Mark D Malamut

Scott Auchmoody

Stephen J Todd

Adam E Brenner

Jennifer Starling

Erik Shawn Hansen

Junxu Li

Christopher J Hackett

Implementations of the disclosure are directed to OCR error correction systems and methods. In some implementations, a method comprises: obtaining, at a computing device, optical character recognition (OCR) text extracted from a document image, the text comprising a token; searching, at the computing device, based on a token bigram determined from the token and a mapping between words in a corpus and a corpus bigram set comprised of unique bigrams from the beginning or ending of the words in the corpus, the corpus for a best word to replace the token; and replacing, at the computing device, the token with the best word.

OCR error correction

Techniques are disclosed for normalized asset valuation and secure storage of information associated with such normalized asset valuation. In one example, a method obtains two or more asset valuation models for a given asset, wherein at least two of the two or more asset valuation models are generated using different asset valuation algorithms, and then converts one or more values in each of the two or more asset valuation models into a common valuation form such that the two or more asset valuation models are converted into two or more normalized valuation models. The common valuation form may be in the form of a cryptocurrency. The method may also combine the two or more normalized valuation models to form a single normalized valuation for the given asset. Still further, the method may store valuation information associated with the converting step in a secure storage system, e.g., a distributed ledger system.

Normalization and secure storage of asset valuation information

A set of valuations is obtained for a set of assets of a given enterprise. The enterprise has a plurality of groups associated therewith, and each of the valuations is computed using a multi-dimensional model configured to generate cross-group impact values with respect to each asset and two or more of the groups associated with the enterprise. One or more storage locations within a storage infrastructure associated with the enterprise to store the assets are determined based on the valuations for the assets. For example, the one or more storage locations may be determined based on a content balancing process that computes an optimal storage location to store each asset in order to minimize overall cost to the enterprise.

Content storage management based on multidimensional valuation models

At least one asset of a given enterprise is identified. The enterprise has a plurality of groups associated therewith. A valuation is computed for the asset using a multi-dimensional model configured to generate cross-group impact values with respect to the asset and two or more of the groups associated with the enterprise. In one example, the multi-dimensional model is in the form of a matrix data structure and the cross-group impact values are off-diagonal entries of the matrix data structure.

Calculating asset value using multi-dimensional models

One example method includes identifying, in a computing system, an aggregate data set that includes both data used by a first user and data used by a second user, examining file metadata associated with the data in the aggregate data set and, based on the examination of the file metadata, determining whether or not any data in the aggregate data set is shared by the first and second users. When a determination is made that some data in the data set is shared by the first and second users, the method includes further determining how much of the data is shared, assigning a strength of connection between the first user and the second user based on the extent to which the first user and second user share data and, based on the strength of connection, taking an action to improve operation and/or configuration of the computing system.

Generating a social graph from file metadata

A data protection ecosystem-based data valuation methodology includes the following steps. One or more of backup data, metadata, and analytics results maintained by a data protection ecosystem are accessed. The backup data, metadata, and analytics results are obtained during the course of the data protection ecosystem providing data backup and recovery functionalities for a data storage environment that stores one or more data sets. A valuation is calculated for at least one of the one or more data sets of the data storage environment based on at least a portion of the accessed backup data, metadata, and analytics results maintained by the data protection ecosystem.

Calculating data value via data protection analytics

One embodiment is related to a method for performing garbage collection at a backup system, comprising: determining an object that is present in a preceding backup but is no longer present in a present backup when or after the present backup is made; associating an indicator indicative of the preceding backup with the object in response to the determination that the object is no longer present in the present backup; determining whether the object is to be deleted from the backup system based on the indicator and a policy during garbage collection; and deleting the object in response to determining that the object is to be deleted.

Garbage collection: timestamp entries and remove reference counts

Embodiments for a method of implementing dynamic subgraphs in a graph database by defining vertices of a subgraph descriptively in such a way that the original intent of the content is preserved while the subgraph itself is changing. A database process preserves a 'view' of the database as a stored query which will define a subgraph members while the subgraph members are modified, and as opposed to a method in which edges are induced from their vertices to be included in the subgraph, by defining the content of a subgraph descriptively using a database view, vertices which are new and should be contained in the subgraph will automatically be included.

Database views for graphs using dynamic subgraphs

Combining a Merkle tree with a graph database by defining a Merkle tree with each node having a hashed value of the metadata for the node and any children of that node, associating non-hashed data with the hashed data for each node, wherein the non-hashed data has an up-pointer from a child node to any of its immediate parent node, and defining bi-directional edges between the nodes of the Merkle tree having a graph database structure to create a reference for the up-pointers associated with each Merkle tree node. The bi-directional edges with up-pointers along with the Merkle tree hash scheme allows efficient traversal of the tree where the hash values indicate non-changed nodes to prevent traversing the database, and further allows efficient path definitions by allowing database processes to walk up edges.

Combining merkle trees in graph databases

Backup objects, native security descriptors that describe access rights to the backup objects, and object metadata are received from platforms where each backup object has corresponding object metadata. Each native security descriptor is transformed into a platform independent security descriptor. A partitioned structure is generated, including by associating each platform independent security descriptor with those object metadata that correspond to the plurality of backup objects for which that platform independent security descriptor describes access rights to.

Secure search

Embodiments for a method of interfacing with a remote application programming interface (API) by defining an execution plan using an interface definition language and a dependency configuration file to generate a constrained directed graph of hierarchically dependent functions of the API, and executing the execution plan using an executing engine that traverses the graph to call the API functions in a defined order and convert data output by a parent function call to input required by a child function call until a terminal vertex is reached that results in directing the resulting in an action such as data to persist and/or affecting the state of a system.

Growing community of inventors

Laguna Niguel, CA, United States of America

Jeffrey Norton