TEMPORAL RULE-BASED FEATURE DEFINITION AND EXTRACTION

Patent №

US 8,538,909

Granted

2013-09-17

Filed 2010

Owner

MICROSOFT CORPORATION

AI components

5

ml · nlp · kr · planning · hardware

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

12972273

A temporal rule-based feature extraction system and method for extracting features from temporal-based rules satisfied by a trace. Once a temporal-based rule is found that is satisfied by the trace, then embodiments of the temporal rule-based feature extraction system and method leverage that rule to either use as a feature or to extract additional features. The extracted feature then is used to characterize the trace. Embodiments of the system include a feature definition module, which defines features based on the temporal-based rules satisfied by a trace, and a similarity measure module, which defines a similarity measure for the defined features. The defined features include both extrinsic features, which are based on extrinsic properties of the rule, and intrinsic features, which are based on intrinsic properties of the rule. The similarity module generates similarity measures that indicate the similarity of two traces.

AI classification

Knowledge representation1.00
Natural language1.00
Machine learning0.99
AI hardware0.88
Planning0.83
Vision0.42
Speech0.00
Evolutionary computation0.00

Ownership

MICROSOFT CORPORATION

assignment · 255910255

Assignors

RANGANTH, VENKATESH-PRASAD, GOYAL, PIYUSH, VALLATHOL, PRADIP HARINDRAN, RAMALINGAM, GANESAN

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC