Raising the baseline for high-precision text classifiers
AssigneeMicrosoft Corporation
InventorsAleksander Kolcz, Wen-tau Yih
Life of this patent
Estimating the term…
Abstract
The claimed subject matter provides systems and/or methods for normalizing document representations for use with Naïve Bayes. The system can include devices and components that determine norms associated with documents by aggregating absolute term weight values associated with the documents, and further ascertain term weights for features associated with the documents, and thereafter divides the term weights for the features associated with the documents with the norms associated with the documents to produce a normalized document representation that can be utilized by arbitrary linear classifiers.
Claims20
A system implemented on a machine that effectuates and facilitates normalization of document representation for use with a Naïve Bayes model, comprising:
- a component that receives a document from an interface, the component determines a norm associated with the document by aggregating absolute term weight values associated with the document and ascertains a term weight for a feature associated with the document, the component divides the term weight for the feature associated with the document with the norm associated with the document to produce a normalized document representation.
A method implemented on a machine that effectuates and facilitates normalization of document representation for use with a Naïve Bayes model, comprising:
- aggregating absolute term weight values associated with a document to determine a norm associated with the document;
- determining a term weight for a feature associated with the document;
- and
- producing a normalized document representation based at least in part on dividing the term weight for the feature associated with the document with the norm associated with the document.
A system that facilitates and effectuates normalization of document representation, comprising:
- means for summing absolute term weight values associated with a document to determine a norm associated with the document;
- means for determining a term weight for a feature associated with the document;
- and
- means for producing a normalized document representation based at least in part on dividing the term weight for the feature associated with the document with the norm associated with the document.
Dependencies are inferred from the claim text.