US 11,734,587 B2
Question answering with time-based weighting
Aaron K. Baughman, Cary, NC (US); Gary F. Diamanti, Wake Forest, NC (US); Mauro Marzorati, Lutz, FL (US); and Elizabeth M. Valletti, Palm Harbor, FL (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Jan. 16, 2020, as Appl. No. 16/744,534.
Application 16/744,534 is a continuation of application No. 14/624,030, filed on Feb. 17, 2015, granted, now 10,572,806.
Prior Publication US 2020/0167675 A1, May 28, 2020
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 5/04 (2023.01); G06N 20/00 (2019.01)
CPC G06N 5/04 (2013.01) [G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method, in a data processing system comprising at least one memory and at least one processor, for question answering using time weighted evidence, the method comprising:
configuring the at least one memory with instructions, which are executed by the at least one processor and configure the at least one processor to implement a question answering system;
receiving, by the question answering system, an input question from a user;
receiving, by the question answering system from the user via a user interface, a time focus for the input question and a curve width, wherein the user interface graphically displays a timeline and a function shape and provides a first control to allow the user to move a center of the function shape along the timeline and a second control to allow the user to change the curve width;
defining, by a time weighting component executing within the question answering system, a weighting function, wherein the weighting function is a bell curve having a peak at the received time focus on a time axis and a width based on the received curve width;
decomposing, by a question decomposition component executing within the question answering system, the input question into one or more queries;
applying, by a hypothesis generation component executing within the question answering system, the one or more queries to a corpus of information to obtain a set of hypothesis evidence, wherein each item of information within the hypothesis evidence has an associated time value;
weighting, by the time weighting component, the set of hypothesis evidence based on the associated time values according to the weighting function to form time weighted evidence;
scoring, by a hypothesis and evidence scoring component executing within the question answering system, a plurality of hypotheses for answering the input question based on the time weighted evidence;
ranking, by a final confidence merging and ranking component executing within the question answering system, the plurality of hypotheses to form a ranked set of candidate answers to the input question; and
outputting, by the question answering system, the ranked set of candidate answers to the user.