Identifying similar digital text volumes
AssigneeGoogle Inc.
InventorsNemanja L. Spasojevic, Guillaume Poncin
Life of this patent
Estimating the term…
Abstract
Relationships between digital text volumes in a corpus are identified by comparing a plurality of digital text volumes in the corpus to identify clusters of similar volumes. For a given cluster, the pages of pairs of volumes within the cluster are compared to determine page similarity data, the page similarity data indicating degrees of similarity between the compared pages. The relationships between pairs of volumes in the clusters are classified based on the similarity data.
Claims20
A computer-implemented method for identifying relationships between a plurality of digital text volumes in a corpus, the method comprising:
- comparing the plurality of digital text volumes in the corpus to produce volume similarity data describing similarities among pairs of digital text volumes;
- identifying a cluster of similar digital text volumes based on the volume similarity data;
- comparing pages of a pair of digital text volumes within the cluster to determine page similarity data indicating degrees of similarity between the compared pages;
- extracting signals from the volume similarity data and the page similarity data that describe a relationship between the pair of digital text volumes, the extracted signals including at least one of:
- a measure of linear page-to-page fit that models volume text uniformly distributed across pages of the pair of digital text volumes;
- and
- a measure of relative consecutive page correlation between pages of the pair of digital text volumes;
- classifying the relationship between the pair of digital text volumes based on the extracted signals;
- and
- storing classification data describing the relationship between the pair of digital text volumes.
A computer system for identifying relationships between a plurality of digital text volumes in a corpus, the system comprising:
- a non-transitory computer-readable storage medium comprising executable computer program code for:
- comparing the plurality of digital text volumes in the corpus to produce volume similarity data describing similarities among pairs of digital text volumes;
- identifying a cluster of similar digital text volumes based on the volume similarity data;
- comparing pages of a pair of digital text volumes within the cluster to determine page similarity data indicating degrees of similarity between the compared pages;
- extracting signals from the volume similarity data and the page similarity data that describe a relationship between the pair of digital text volumes, the extracted signals including at least one of:
- a measure of linear page-to-page fit that models volume text uniformly distributed across pages of the pair of digital text volumes;
- and
- a measure of relative consecutive page correlation between pages of the pair of digital text volumes;
- classifying the relationship between the pair of digital text volumes based on the extracted signals;
- and
- storing classification data describing the relationship between the pair of digital text volumes;
- and
- a processor for executing the computer program code.
A non-transitory computer-readable storage medium storing executable computer program instructions for identifying relationships between a plurality of digital text volumes in a corpus, the computer program instructions comprising instructions for:
- comparing the plurality of digital text volumes in the corpus to produce volume similarity data describing similarities among pairs of digital text volumes;
- identifying a cluster of similar digital text volumes based on the volume similarity data;
- comparing pages of a pair of digital text volumes within the cluster to determine page similarity data indicating degrees of similarity between the compared pages;
- extracting signals from the volume similarity data and the page similarity data that describe a relationship between the pair of digital text volumes, the extracted signals including at least one of:
- a measure of linear page-to-page fit that models text uniformly distributed across pages of the pair of digital text volumes and
- a measure of relative consecutive page correlation between pages of the pair of digital text volumes;
- classifying the relationship between the pair of digital text volumes based on the extracted signals;
- and
- storing classification data describing the relationship between the pair of digital text volumes.
Dependencies are inferred from the claim text.