US 12,189,669 B2
Extracting query-related temporal information from unstructured text documents
Udit Sharma, Gurgaon (IN); Hima Prasad Karanam, Bangalore (IN); Shajith Ikbal Mohamed, Chennai (IN); Sumit Neelam, Bangalore (IN); Santosh Srivastava, New Delhi (IN); and L. Venkata Subramaniam, New Delhi (IN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Dec. 6, 2021, as Appl. No. 17/543,127.
Prior Publication US 2023/0177076 A1, Jun. 8, 2023
Int. Cl. G06F 16/33 (2019.01); G06F 17/10 (2006.01)
CPC G06F 16/334 (2019.01) [G06F 17/10 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
obtaining at least one user query;
converting at least a portion of the at least one user query into one or more logic form representations which capture one or more temporal constraints indicated in one or more portions of content within the at least one user query, wherein converting comprises processing the at least a portion of the at least one user query using one or more neural networks;
mapping at least a portion of the one or more logic form representations to unstructured text data stored in one or more portions of at least one source of unstructured text data, wherein the at least one source of unstructured text data comprises at least one query-relevant knowledge base;
extracting one or more items of temporal information, specific to the at least one user query, from one or more portions of the at least one source of unstructured text data based at least in part on the mapping;
generating at least one response to the at least one user query based at least in part on the one or more extracted items of temporal information;
wherein at least portions of results of the obtaining step, the converting step, the mapping step, the extracting step, and the generating step are stored in one or more knowledge bases; and
performing one or more automated actions based at least in part on the at least one generated response, wherein performing the one or more automated actions comprises (i) transmitting the at least one generated response to at least one user corresponding to the at least one user query and (ii) automatically training at least a portion of the one or more neural networks using at least a portion of the at least one user query, comprising the at least a portion of the one or more logic form representations, and at least a portion of the at least one generated response;
wherein the method is carried out by at least one computing device.