US 11,755,572 B1
Systems, methods, and apparatus for natural language queries and data visualizations
Bibek Mishra, Bhubaneswar (IN); Anubha Gupta, Hartford, CT (US); and Reena Samuel, Monroe, CT (US)
Assigned to The Travelers Indemnity Company, Hartford, CT (US)
Filed by The Travelers Indemnity Company, Hartford, CT (US)
Filed on Jun. 26, 2020, as Appl. No. 16/913,936.
Int. Cl. G06F 16/2452 (2019.01); G06F 16/2457 (2019.01); G06F 16/242 (2019.01); G06F 40/284 (2020.01)
CPC G06F 16/24522 (2019.01) [G06F 16/2428 (2019.01); G06F 16/2448 (2019.01); G06F 16/24573 (2019.01); G06F 40/284 (2020.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method for converting a natural language query into a structured database query, the computer-implemented method comprising:
receiving, by a dashboard controller system, natural language query text of a user, wherein the natural language query text is indicative of a request of the user for information from a database system;
transforming, by the dashboard controller system, each word of the natural language query text to a respective base lemma, thereby generating a set of transformed words corresponding to the natural language query text;
comparing, by the dashboard controller system, each transformed word of the set of transformed words to a database corpus of stored data elements to determine whether the transformed word has a respective matching data element in the database corpus, wherein the database corpus comprises a first data ecosystem;
based on the comparing, identifying a first unmatched transformed word of the set of transformed words that does not have a respective matching data element in the database corpus;
requesting, by the dashboard controller system, and from the user, a user input to assist in correlating the first unmatched transformed word;
receiving, by the dashboard controller system, and in response to the requesting, a user response, wherein the user response comprises an indication that the first unmatched transformed word matches a second data element of a second data ecosystem;
generating, by the dashboard controller system, a query builder collection based on the user response and the set of transformed words;
generating, by the dashboard controller system, a structured data query based on the query builder collection;
executing, by the dashboard controller system, the structured data query against at least one data table of the database system to generate a data query result;
selecting, by the dashboard controller system, a data visualization for the user;
generating, by the dashboard controller system, a display of the data query result via a personalized data query dashboard based on the selected data visualization for the user;
transmitting, by the dashboard controller system to the user, a request for feedback information about the generated display of the data query result;
receiving, by the dashboard controller system from the user, feedback information about the generated display of the data query result;
and updating at least one data module of the database system based on the feedback information.