Cross-validation based code feature tagging
AssigneeInternational Business Machines Corporation
InventorsYi-Min Chee, Hao Chen, Ya Bin Dang, Shao Chun Li, Guang Tai Liang, Li Jun Mei
Life of this patent
Estimating the term…
Abstract
Method and systems for tagging code include organizing code units in a hierarchical structure that associates the code units with respective description units. The hierarchical structure is filtered to remove un-meaningful description units from the hierarchical structure. Feature tags are generated based on a weighted sum of term frequencies in the description units of the hierarchical structure to associate features with respective code units.
Claims20
A method for tagging code, comprising:
- organizing code units in a hierarchical structure that associates the code units with respective description units;
- filtering the hierarchical structure to remove un-meaningful description units from the hierarchical structure;
- and
- generating feature tags, using a processor, based on a weighted sum of term frequencies in the description units of the hierarchical structure to associate features with respective code units.
A computer readable storage medium comprising a computer readable program for tagging code, wherein the computer readable program when executed on a computer causes the computer to perform the steps of:
- organizing code units in a hierarchical structure that associates the code units with respective description units;
- filtering the hierarchical structure to remove un-meaningful description units from the hierarchical structure;
- and
- generating feature tags, using a processor, based on a weighted sum of term frequencies in the description units of the hierarchical structure to associate features with respective code units.
A feature tagging system, comprising:
- a hierarchical structure generation module configured to organize code units in a hierarchical structure that associates code units with respective description units;
- a refinement module configured to filter the hierarchical structure to remove un-meaningful description units from the hierarchical structure;
- and
- a feature tag module comprising a processor configured to generate feature tags based on a weighted sum of term frequencies in the description units of the filtered hierarchical structure to associate features with respective code units.
Dependencies are inferred from the claim text.