Filtering of large sets of data
AssigneeMicrosoft Technology Licensing, LLC
InventorsEran Gat, Moshe Goldstine, Elad Yom-Tov
Life of this patent
Estimating the term…
Abstract
A computer-implemented filtering method comprising: receiving a set of propositions, each comprising a respective linguistic description expressed by a respective user; forming an input matrix of lexical components vs. propositions, wherein rows in the input matrix represent the propositions and columns in the input matrix represent lexical components in the respective descriptions; performing a matrix factorization on the input matrix to reveal latent clusters of the lexical components and/or propositions; filtering the set of propositions based on one or more of the latent clusters; and outputting a result of the filtering to a target user.
Claims20
A computer-implemented filtering method comprising:
- receiving a set of propositions, each comprising a respective description expressed by a respective user;
- forming an input matrix of lexical components versus propositions, wherein rows in the input matrix represent the propositions and columns in the input matrix represent lexical components in the respective descriptions;
- performing a matrix factorization on the input matrix to reveal latent clusters of at least one of the lexical components or propositions;
- filtering said set of propositions based on one or more of said latent clusters, wherein said filtering comprises:
- evaluating a measure of vector similarity between a target proposition vector and each of a plurality of said latent clusters of propositions, the target proposition vector comprising a vector of propositions expressed by a target user;
- and
- selecting a latent cluster of propositions based on said evaluating;
- and
- outputting a result of the filtering to the target user.
A computer program product comprising code embodied on computer-readable storage and configured so as when run on one or more processing units to perform operations of:
- receiving a set of propositions, each comprising a respective description expressed by a respective user;
- forming an input matrix of lexical components versus propositions, wherein rows in the input matrix represent the propositions and columns in the input matrix represent lexical components in the respective descriptions;
- performing a matrix factorization on the input matrix to reveal latent clusters of at least one of the lexical components or propositions;
- filtering said set of propositions based on one or more of said latent clusters, wherein said filtering comprises:
- evaluating a measure of vector similarity between a target proposition vector and each of a plurality of said latent clusters of propositions, the target proposition vector comprising a vector of propositions expressed by a target user;
- and
- selecting a latent cluster of propositions based on said evaluating;
- and
- outputting a result of the filtering to a target user.
A recommendation system comprising:
- storage comprising one or more memory units, and
- processing apparatus comprising one or more processing units;
- wherein the storage stores code and the processing apparatus is arranged to run said code, the code being configured so as when thus run on the processing apparatus to perform operations of:
- receiving a set of propositions, each comprising a respective description expressed by a respective user;
- forming an input matrix of lexical components versus propositions, wherein rows in the input matrix represent the propositions and columns in the input matrix represent lexical components in the respective descriptions;
- performing a matrix factorization on the input matrix to reveal latent clusters of at least one of the lexical components or propositions;
- filtering said set of propositions based on one or more of said latent clusters, wherein said filtering comprises:
- evaluating a measure of vector similarity between a target proposition vector and each of a plurality of said latent clusters of propositions, the target proposition vector comprising a vector of propositions expressed by a target user;
- and
- selecting a latent cluster of propositions based on said evaluating;
- and
- outputting a result of the filtering to a target user.
Dependencies are inferred from the claim text.