US 12,248,472 B2
Systems and methods for dynamic query prediction and optimization
Denise Rogers, West Hartford, CT (US); Yulanda Henry, Hartford, CT (US); David P. Therrien, Wallingford, CT (US); Rebecca Peralta, West Hartford, CT (US); Adam J. Rychlik, Bristol, CT (US); and Bibek Mishra, Manchester, CT (US)
Assigned to The Travelers Indemnity Company, Hartford, CT (US)
Filed by The Travelers Indemnity Company, Hartford, CT (US)
Filed on Jan. 27, 2023, as Appl. No. 18/160,490.
Application 18/160,490 is a continuation of application No. 16/588,356, filed on Sep. 30, 2019, granted, now 11,593,370.
Prior Publication US 2023/0169078 A1, Jun. 1, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/00 (2019.01); G06F 16/215 (2019.01); G06F 16/242 (2019.01); G06F 16/2453 (2019.01); G06N 5/04 (2023.01); G06N 20/00 (2019.01)
CPC G06F 16/24542 (2019.01) [G06F 16/215 (2019.01); G06F 16/2425 (2019.01); G06N 5/04 (2013.01); G06N 20/00 (2019.01)] 26 Claims
OG exemplary drawing
 
1. A controller device for query prediction, comprising:
a processor; and
a computer-readable medium storing instructions that when executed by the processor direct the processor to:
receive, from a user query device, a structured query language (SQL) statement for tuning;
automatically transform the SQL statement to generate a cleaned SQL statement, wherein the transforming comprises at least one of (i) identifying and converting any date within the SQL statement to a standardized format and (ii) identifying and removing at least one word from the SQL statement;
automatically generate at least one query feature string based on the cleaned SQL statement, each query feature string corresponding to a respective SQL feature of the cleaned SQL statement;
automatically generate, for each query feature string, a respective feature count vector comprising an indication of the corresponding SQL feature and an indication of a number of occurrences of the corresponding SQL feature in the cleaned SQL statement, thereby generating a dataset of feature count vectors;
automatically assemble, based on the dataset of feature count vectors, an aggregate feature vector;
access a dataset of historical aggregate feature vectors;
automatically generate an SQL statement performance prediction for the SQL statement by testing the aggregate feature vector against the dataset of historical aggregate feature vectors,
wherein generating the SQL statement performance prediction for the SQL statement does not comprise executing the SQL statement; and
output the SQL statement performance prediction.