Managing questioning in a question and answer system
A system, a method, and a computer program product for managing questioning in a Question and Answering (QA) system are disclosed. An input question received by the QA system is analyzed to determine the set of answer-types. A set of answer-types is determined. Prompting for clarifying data associated with the input question happens. A set of candidate answers of the answer-type is determined. The clarifying data is used to determine the set of candidate answers. At least one candidate answer of the set of candidate answers of the answer-type is provided.
1. A computer-implemented method for managing questioning in a Question and Answering (QA) system, the method comprising:
receiving, by the QA system, an input question from a user, wherein the input question is associated with a particular product, and wherein the product is a physical product having multiple different generations;
determining, by performing natural language processing on the input question, that the input question calls for a temporally-dependent answer, wherein the determining that the input question calls for a temporally-dependent answer comprises:
identifying, based on the natural language processing on the input question, a major defining characteristic for the input question;
determining that the major defining characteristic is time-dependent, wherein the major defining characteristic is a time span dependent feature of the product, and wherein the time span dependent feature of the product is different for each different generation; and
identifying a plurality of potential values of the major defining characteristic, wherein each identified potential value is associated with a different time span;
prompting, in response to the determining that the input question calls for a temporally-dependent answer, the user to provide temporally-related data to clarify the input question, wherein the prompting the user to provide temporally-related data to clarify the input question comprises:
prompting the user to indicate which of the plurality of values most closely pertains to the user's basis for the input question;
receiving, from the user and in response to the prompting, the temporally-related data;
determining, using the received temporally-related data and the input question, a set of candidate answers to the input question; and
providing at least one candidate answer of the set of candidate answers to the user.
2. A computer-implemented method for managing questioning in a Question and Answering (QA) system, the method comprising:
receiving, by the QA system, an input question from a user, wherein the input question is associated with a particular product, and wherein the product is a physical product having multiple different generations;
identifying, based on performing natural language processing on the input question, a major defining characteristic for the input question, wherein the major defining characteristic is a time span dependent feature of the product, and wherein the time span dependent feature of the product is different for each different generation;
identifying a plurality of potential values of the major defining characteristic;
prompting the user to indicate which of the plurality of potential values most closely pertains to the user's basis for the input question;
receiving, from the user and in response to the prompting, an indication of which of the plurality of potential values most closely pertains to the user's basis for the input question;
determining, using the indication from the user and the input question, a set of candidate answers to the input question; and
providing at least one candidate answer of the set of candidate answers to the user.