Scalable string matching as a component for unsupervised learning in semantic meta-model development

An apparatus for use in string analysis, comprising:

• a feature module, executable by a processor of the computer-based string analysis tool, operable to represent each of a plurality of target strings as a plurality of features;
• an indexing module, executable by a processor of the computer-based string analysis tool, operable to compile an index comprising a mapping, wherein the mapping correlates each target feature to each one of the plurality of target strings containing the target feature;
• a similarity metric module, executable by a processor of the computer-based string analysis tool, operable to calculate a similarity metric between a source string and each of a selected portion of the plurality of target strings having at least one common feature to the source string, wherein the similarity metric comprises a quantitative value regarding the commonality of features between the source string and each of the selected portion of the plurality of target strings;
• and
• an optimization module, executable by a processor of the computer-based string analysis tool, operable to determine the selected portion of the plurality of target strings for which a similarity metric is calculated based at least partially on elimination of features from consideration based on a minimum similarity threshold.

A non-transitory storage medium for storing instructions for execution by a processor of a computer-based string analysis tool, wherein the instructions, when executed by the processor, cause the processor to:

• identify a plurality of target strings;
• represent each of the plurality of target strings as one or more target features;
• compile an index corresponding to the target features, wherein the index comprises a mapping for each target feature to each of the plurality of target strings containing the target feature;
• receive a source string;
• represent the source string as a plurality of source features;
• determine a portion of the target strings for which a similarity metric is to be calculated at least partially based on the index, wherein the portion of the target strings comprises less than the entirety of the target strings;
• and
• calculate, using the computer-based comparison tool, a similarity metric for the source string relative to only the portion of the plurality of target strings, wherein the similarity metric is at least partially based on commonality between the target features and the source features.