Systems and method for generating a structured report from unstructured data

A system for refining an extraction rule, comprising:

• at least one unstructured document source;
• a server configured to:
• receive at least one unstructured document;
• apply at least one extraction rule to the at least one unstructured document to identify data that is relevant to a query, the query configured to identify at least one target term;
• generate, using a recommendation engine, at least one suggested filter recommendation to modify the at least one extraction rule to refine the data identified as relevant to the query, based on at least one path of a decision tree trained with data previously identified as relevant to the query,
• wherein the data previously identified as relevant to the query includes one or more positive sentences and one or more negative sentences,
• wherein the one or more positive sentences include the at least one target term,
• wherein the one or more negative sentences exclude the at least one target term, and
• wherein generating the at least one suggested filter recommendation includes:
• extracting the at least one path of the decision tree;
• and
• converting the extracted at least one path into at least one Boolean expression;
• display the at least one Boolean expression as the at least one suggested filter recommendation to a user;
• receive a selection of one or more of the at least one suggested filter recommendation by the user;
• and
• modify the at least one extraction rule based on the selected one or more suggested filter recommendation.

A computer-based tool for refining an extraction rule, the computer-based tool including non-transitory computer readable media having stored thereon computer code which, when executed by a processor, causes a computing device to perform operations comprising:

• receiving at least one input file containing unstructured data;
• applying at least one extraction rule to the at least one input file to identify data that is relevant to a query, the query configured to identify at least one target term;
• generating, by a recommendation engine, at least one suggested filter recommendation to modify the at least one extraction rule to refine the data identified as relevant to the query, based on at least one path of a decision tree trained with data previously identified as relevant to the query,
• wherein the data previously identified as relevant to the query includes one or more positive sentences and one or more negative sentences,
• wherein the one or more positive sentences include the at least one target term,
• wherein the one or more negative sentences exclude the at least one target term, and
• wherein generating the at least one suggested filter recommendation includes:
• extracting the at least one path of the decision tree;
• and
• converting the extracted at least one path into at least one Boolean expression;
• displaying the at least one Boolean expression as the at least one suggested filter recommendation to a user;
• receiving a selection of one or more of the at least one suggested filter recommendation by the user;
• and
• modifying the at least one extraction rule based on the selected one or more suggested filter recommendation.