Systems and methods for configuring system memory for extraction of latent information from big data
AssigneeReservoir Labs, Inc.
InventorsJames Ezick, Thomas Henretty, Richard A. Lethin
Life of this patent
Estimating the term…
Abstract
A system for extracting latent information from data includes obtaining or generating components of the data, where the data components include scores indicating how the component relates to the data. Memory is allocated for the components and the components are stored in the allocated memory. The components are then transformed into documents using a suitable transformation function, and the documents are analyzed using natural language processing, to extract latent information contained in the data.
Claims51
A method for configuring memory of a computing system for extraction of latent information from data comprising a first dataset represented in a first index space, the method comprising:
- obtaining a plurality of components comprising a set of components of the first dataset, wherein each component in the plurality of components comprises a plurality of scores associated with a second index space that is different from the first index space, wherein a respective label is associated with each index in the second index space;
- allocating in the memory, for at least one component of the plurality of components, a block of memory that is designated as a vector and has a size at least equal to a total number of labels corresponding to non-zero scores in the at least one component;
- for each component of the plurality of components, storing in the respective vector, scores corresponding to a plurality of labels associated with the second index space, wherein a common label sequence is used for storing the scores in each vector;
- generating a plurality of documents, wherein each document corresponds to a respective component and is obtained by applying a transform function to the vector corresponding to that component;
- and
- analyzing the plurality of documents using a processor adapted to perform natural language processing (NLP), wherein analysis from the NLP is expressed using one or more labels associated with the second index space, thereby describing latent information contained in the data.
A system for configuring memory used for extraction of latent information from data comprising a first dataset represented in a first index space, the system comprising:
- a first processor;
- a first memory in electrical communication with the first processor, the first memory comprising instructions which, when executed by a processing unit comprising at least one of the first processor and a second processor, and in electronic communication with a memory module comprising at least one of the first memory and a second memory, program the processing unit to:
- obtain a plurality of components comprising a set of components of the first dataset, wherein each component in the plurality of components comprises a plurality of scores associated with a second index space that is different from the first index space, wherein a respective label is associated with each index in the second index space;
- allocate in the memory module, for at least one component of the plurality of components, a block of memory that is designated as a vector and has a size at least equal to a total number of labels corresponding to non-zero scores in the at least one component;
- store in the respective vector for each component of the plurality of components, scores corresponding to a plurality of labels associated with the second index space, wherein a common label sequence is used for storing the scores in each vector;
- and
- generate a plurality of documents, wherein each document corresponds to a respective component and is obtained by applying a transform function to the vector corresponding to that component.
Dependencies are inferred from the claim text.