Back to results
US 2026/0289361 A1Published application
Database utilizing spatial probability models for data compression
OwnerAdam Jeffriesinventor
Life of this patent
Estimating the term…
Abstract
A method, article comprising machine-readable instructions and apparatus that processes data systems for encoding, decoding, pattern recognition/matching and data generation is disclosed. State subsets of a data system are identified for the efficient processing of data based, at least in part, on the data system's systemic characteristics.
Claims18
A system comprising:
- a processing device;
- computerized systema non-transitory, tangible medium that stores instructions that, when executed by the processing device, cause the system to perform operations comprising:
- receiving, by a data processor comprising hardware, a data system comprising an ordered sequence of data units that manifest, as a whole, a state comprised by a state superset, the data system corresponding to at least one of:
- an encoded message comprising symbols, a compressed message comprising bits, a document comprising graphemes and/or alphanumeric symbols, a digital picture comprising pixels, a waveform file comprising samples, a video file comprising picture frames, a data file comprising bytes, or an executable comprising machine code symbols;
- applying a sequence function to the ordered sequence of data units to generate a monotonic output sequence;
- determining a correspondence between members of the monotonic output sequence and elements of a space of a spatial probability model, wherein the spatial probability model reduces memory utilization and/or utilization of computational resources by incorporating a sequence of probability distributions that vary along intervals in a time dimension comprised by the space of the spatial probability model, the time dimension corresponding to at least a length of the monotonic output sequence;
- utilizing the spatial probability model to determine a shape of at least one of:
- the monotonic sequence and/or a boundary comprising the monotonic sequence, the shape based, at least in part, on the correspondence between members of the monotonic output sequence and elements of the space of the spatial probability model;
- and
- determining a similarity between the shape and at least one second shape based, at least in part, on at least one second spatial probability model by performing at least one of the following:
- fitting and/or matching the shape to at least one second shape corresponding to elements comprised by at least one second spatial probability model, and/or determining the shape fits, matches, is similar to and/or is enclosed by probability constraints corresponding to at least one second spatial probability model,
- whereby matching the shape identifies a systemic characteristic and/or pattern of the data system.
9 more elements
A system comprising:
- a processing device;
- computerized systema non-transitory, tangible medium that stores instructions that when executed by the processing device cause the system to perform operations comprising at least:
- receiving, by a data processor comprising hardware, a data system comprising an ordered sequence of data units that manifest, as a whole, a state comprised by a state superset, the data system corresponding to at least one of:
- an encoded message comprising symbols, a compressed message comprising bits, a document comprising graphemes and/or alphanumeric symbols, a digital picture comprising pixels, a waveform file comprising samples, a video file comprising picture frames, a data file comprising bytes, or an executable comprising machine code symbols;
- applying a sequence function to the ordered sequence to generate a monotonic output sequence;
- determining a correspondence between members of the monotonic output sequence and elements of a space of a spatial probability model, wherein the spatial probability model reduces memory utilization and/or utilization of computational resources by incorporating a sequence of probability distributions that vary along intervals in the time dimension comprised by the space of the spatial probability model, the time dimension corresponding to at least a length of the monotonic output sequence;
- determining a probability value for possible states represented by the elements of the space using at least one of:
- a continuous probability distribution function, a discrete probability distribution function, and/or a probability mapping function;
- and
- performing at least one of:
- selecting states and/or sequences that correspond to elements associated with a higher probability than other elements;
- generating a generated monotonic sequence where its members correspond to the most likely state, and/or;
- generating a generated data system using constraints of the spatial probability model as guides for the generation, wherein the generated data system comprises data units manifesting states comprised by a positive probability subspace of the spatial probability model.
10 more elements
A method for constructing a spatial probability model that relatively reduces memory utilization, computational resources and information distortion by comprising probability distributions that vary along intervals in a time dimension corresponding to a length of a monotonic output sequence to determine a probability of a state comprised by a state superset of at least one of the following:
- a data unit, the monotonic output sequence, a data system or a systemic characteristic of the data system, the method comprising:
- receiving, by a data processing device, at least one data system comprising an ordered sequence of data units;
- applying a sequence function to the ordered sequence to generate a monotonic output sequence;
- determining a correspondence between members of the monotonic output sequence and elements of a space comprised by the spatial probability model;
- recording a number of correspondences between the members and the elements of the space;
- generating probability values for possible states represented by elements within a positive probability subspace comprised by the space of the spatial probability model using at least one of:
- a continuous probability distribution function that generates a probability gradient based, at least in part, on the recorded number of correspondences,
- a discrete probability distribution function, and/or
- a probability mapping function that assigns probability values to elements of the space based, at least in part, on the recorded number of correspondences;
- and
- storing, by the data processor, the probability values in a storage medium, wherein the spatial probability model reduces memory utilization and/or utilization of computational resources when utilizing the spatial probability model to perform at least one of:
- matching systemic characteristic patterns of an incoming data system with systemic characteristics described by the spatial probability model,
- forecasting by a forecaster a forecasted data system, the incoming data system and the forecasted data system corresponding to at least one of:
- an encoded message comprising symbols, a document comprising graphemes or alphanumeric symbols, a digital picture comprising pixels, a waveform file comprising samples, a video file comprising picture frames, a digital scene comprising points, polygons or voxels, a data file comprising bytes, a pseudorandom file comprising data with pseudorandom characteristics, or a forecasted ordered sequence of data units corresponding to predicted data associated with a likelihood of occurring at the future interval or the previous interval comprised by the time dimension.
11 more elements
Dependencies are inferred from the claim text.