System and method for an expert question answer system from a dynamic corpus

A system comprising:

• at least one computing device configured to update an expert corpus set, by performing actions including:
• obtaining a query from a first user;
• obtaining a first raw data source;
• determining a first relevance score for the first raw data source with respect to the query, by performing actions including:
• creating a first vector of statistical variables for the query using at least one natural language processing (NLP) socket, the statistical variables of the first vector having category types;
• creating a second vector of statistical variables for the first raw data source, the statistical variables for the first raw data source having category types that are the same as the category types of the statistical variables for the query;
• and
• generating a hypothesis regarding the relevance of the first raw data source with respect to the query;
• testing the hypothesis by comparing each statistical variable for the query to each same statistical variable for the first raw data source;
• and
• calculating a gradient between the first vector and the second vector to determine the first relevance score;
• and
• updating the expert corpus set by ingesting the first raw data source into the expert corpus in response to determining the first relevance score exceeding a first threshold.

A computer program product comprising program code stored on a computer-readable storage medium, which when executed by at least one computing device, enables the at least one computing device to implement a method of updating an expert corpus set by performing actions including:

• obtaining a query from a first user;
• obtaining a first raw data source;
• determining a first relevance score for the first raw data source with respect to the query, by performing actions including:
• creating a first vector of statistical variables for the query using at least one natural language processing (NLP) socket, the statistical variables of the first vector having category types including at least one of lexical answer type, sentence focus frequencies or entity type;
• creating a second vector of statistical variables for the first raw data source, the statistical variables for the first raw data source having category types that are the same as the category types of the statistical variables for the query;
• generating a hypothesis regarding the relevance of the first raw data source with respect to the query;
• testing the hypothesis by comparing each statistical variable for the query to each same statistical variable for the first raw data source;
• calculating a gradient between the first vector and the second vector to determine the first relevance score;
• and
• updating the expert corpus set by ingesting the first raw data source into the expert corpus in response to determining the first relevance score exceeding a first threshold.