Back to results
US 11,106,878 B2Granted
Generating hypotheses in data sets
AssigneeGeorgetown University
InventorsOphir Frieder, David Hartley
Life of this patent
Estimating the term…
Abstract
A method for generating hypotheses in a corpus of data comprises selecting a form of ontology; coding the corpus of data based on the form of the ontology; generating ontology space based on coding results and the ontology; transforming the ontology space into a hypothesis space by grouping hypotheses; weighing hypotheses included in the hypothesis space; and applying a science-based optimization algorithm configured to model a science-based treatment of the weighted hypotheses.
Claims17
A method of identifying hypotheses in a corpus of data, the method comprising:
- receiving an ontology by one or more computers, wherein the ontology includes fields and a plurality of choices for each of the fields such that the ontology includes a plurality of ontology vectors, each ontology vector including one choice for each of the fields;
- receiving the corpus of data by the one or more computers;
- populating the ontology, by the one or more computers, by detecting data from the corpus of data that correspond to each of the ontology vectors;
- plotting the populated ontology vectors in an ontology space by the one or more computers, each dimension of the ontology space being associated with one or more of the fields of the ontology;
- transforming the plotted ontology space into a hypothesis space, by the one or more computers, by grouping the plotted ontology vectors that describe similar and/or related concepts into neighborhoods, wherein each neighborhood represents a hypothesis;
- weighing each of the plotted hypotheses by the one or more computers;
- plotting a hypothesis surface or population in the hypothesis space, by the one or more computers, such that the hypothesis surface has troughs or the population has members, wherein the depth of each trough or fitness of each population member is proportional to the weight of each hypothesis in the hypothesis space;
- and
- sorting the hypotheses plotted in the hypothesis space, by the one or more computers, by applying an optimization algorithm to find the lowest troughs in the hypothesis surface or fittest population members in the population, wherein the optimization algorithm comprises one of a simulated annealing algorithm, a random walk algorithm, a Monte Carlo-based algorithm, or a genetic algorithm.
6 more elements
A system for identifying hypotheses in a corpus of data, the method comprising:
- non-transitory computer readable storage media that stores the corpus of data;
- a content server that:
- receives an ontology, the ontology including fields and a plurality of choices for each of the fields such that the ontology includes a plurality of ontology vectors, each ontology vector including one choice for each of the fields;
- populates the ontology by detecting data from the corpus of data that correspond to each of the ontology vectors;
- plots the populated ontology vectors in an ontology space, each dimension of the ontology space being associated with one or more of the fields of the ontology;
- transforms the plotted ontology space into a hypothesis space by grouping the plotted ontology vectors that describe similar and/or related concepts into neighborhoods, wherein each neighborhood represents a hypothesis;
- weighs each of the hypotheses plotted in the hypothesis space;
- plots a hypothesis surface or population in the hypothesis space such that the hypothesis surface has troughs or the population has members, wherein the depth of each trough or fitness of each population member is proportional to the weight of each hypothesis in the hypothesis space;
- and
- sorts the hypotheses plotted in the hypothesis space by applying an optimization algorithm to find the lowest troughs in the hypothesis surface or fittest of population members in the population, wherein the optimization algorithm comprises one of a simulated annealing algorithm, a random walk algorithm, a Monte Carlo-based algorithm, or a genetic algorithm.
7 more elements
Dependencies are inferred from the claim text.