System and method of machine-aided information extraction rule development

A computer-implemented method of automatic rule generation, comprising:

• (a) generating a rule from a sample;
• (b) generating a relaxed rule from the rule and, if the relaxed rule matches a stored rule, returning the relaxed rule as an appropriate extraction rule;
• (c) generating a reverse index of a corpus;
• (d) generating text segments by applying the relaxed rule to the reverse index;
• (e) generating modified text segments from the text segments and the relaxed rule;
• (f) performing a candidate generation process on the modified text segments;
• (g) if the candidate generation process generates no candidates, performing again (b) using a further relaxed rule as the relaxed rule;
• (h) evaluating, by a user, the candidate;
• (i) providing, by the user, the candidate as an additional sample;
• and
• (j) performing again (a) using the sample and the additional sample to generate another rule to use as the rule.

A computer program product, comprising a computer readable medium having a computer readable program code embodied therein, that when executed by a data processing apparatus, implements an automatic rule generation system, comprising:

• a rule generation module that receives a sample and generates a rule from the sample;
• a rule relaxation module that generates a relaxed rule from the rule;
• a rule testing module that generates a reverse index from a corpus, applies the relaxed rule to the reverse index, and generates text segments;
• an information extraction module that generates modified text segments from the relaxed rule and the text segments;
• and
• a candidate suggestion module that performs a candidate generation process, wherein if the candidate generation process generates no candidates, the candidate suggestion module signals the rule relaxation module to generate a further relaxed rule to use as the relaxed rule.