Processing data for use in a cognitive insights platform

A method, comprising:

• receiving, by a device, issue resolution information associated with a cognitive model, the issue resolution information including an item of issue resolution information that describes an issue and a resolution to the issue;
• identifying, by the device, a plurality of domain hierarchies associated with one or more domains that may be related to the issue resolution information, wherein the one or more domains include a category or field of knowledge to which the issue corresponds, and wherein the plurality of domain hierarchies are associated with different issue resolution information;
• automatically assigning, by the device and without user intervention, the item of issue resolution information to a particular domain hierarchy of the plurality of domain hierarchies;
• determining, by the device, a confidence score indicating a degree of confidence associated with assigning the item of issue resolution information to the particular domain hierarchy;
• providing, by the device and to a first user, an interface to review or modify the particular domain hierarchy, the first user being an administrator associated with the device, the first user to manually review or modify the particular domain hierarchy automatically selected from the plurality of domain hierarchies, another domain hierarchy, of the plurality of domain hierarchies, being provided based on the confidence score not satisfying a threshold, and the first user to select the particular domain hierarchy via at least one of:
• a drop down menu, or
• a radio button;
• receiving, by the device, user input associated with assigning the item of issue resolution information to the particular domain hierarchy or other domain hierarchy, the user input associated with assigning the item of issue resolution information to the particular domain hierarchy or other domain hierarchy being received from the first user;
• determining, by the device, one or more questions corresponding to the item of issue resolution information and an answer associated with the one or more questions, the determining of the one or more questions and the answer being based on user input by a second user that is different from the first user, the second user being prevented from editing or modifying the assigning of the item of issue resolution information to the particular domain hierarchy or other domain hierarchy;
• approving, by the device, the one or more questions and the answer, the approving of the one or more questions and the answer being associated with a third user that is different from the first user and the second user;
• generating, by the device and based on approving the one or more questions and the answer, one or more question/answer (QA) pairs, each QA pair, of the one or more QA pairs, being associated with a question, of the one or more questions, and the answer;
• creating, by the device, a data corpus including the one or more QA pairs;
• causing, by the device, the cognitive model to be trained based on a portion of the data corpus;
• determining, by the device, a performance metric based on an output of the cognitive model;
• determining, by the device, whether the performance metric satisfies a threshold corresponding to the performance metric;
• and
• selectively:
• validating, by the device, the training of the cognitive model when the performance metric satisfies the threshold corresponding to the performance metric, or
• causing, by the device, the cognitive model to be retrained when the performance metric does not satisfy the threshold corresponding to the performance metric.

A non-transitory computer-readable medium storing instructions, the instructions comprising:

• one or more instructions that, when executed by one or more processors of a device, cause the one or more processors to:
• receive issue resolution information associated with a cognitive model, the issue resolution information including an item of issue resolution information that describes an issue and a resolution corresponding to the issue;
• identify a plurality of domain hierarchies associated with one or more domains that may be related to the issue resolution information, wherein the one or more domains include a category or field of knowledge to which the issue corresponds, and wherein the plurality of domain hierarchies is associated with different issue resolution information;
• automatically assign, without user intervention, the item of issue resolution information to a particular domain hierarchy, of the plurality of domain hierarchies;
• determine a confidence score indicating a degree of confidence associated with assigning the item of issue resolution information to the particular domain hierarchy;
• provide, to a first user, an interface to review or modify the particular domain hierarchy, the first user being an administrator associated with the device, the first user to review or modify the particular domain hierarchy automatically selected from the plurality of domain hierarchies, and another domain hierarchy, of the plurality of domain hierarchies, being provided based on the confidence score not satisfying a threshold;
• receive user input associated with assigning the item of issue resolution information to the particular domain hierarchy or the other domain hierarchy, the user input associated with assigning the item of issue resolution information to the particular domain hierarchy or the other domain hierarchy being received from the first user;
• generate, based on assigning the item of issue resolution information to the particular domain hierarchy or the other domain hierarchy, a question and an answer corresponding to the item of issue resolution information, the generating of the question and the answer being based on user input by a second user that is different from the first user, and the second user being prevented from editing or modifying the assigning of the item of issue resolution information to the particular domain hierarchy or the other domain hierarchy;
• approve, based on generating the question and the answer, the question and the answer corresponding to the item of issue resolution information;
• generate, based on approving the question and the answer, a question/answer (QA) pair corresponding to the question and the answer;
• create a data corpus including the QA pair;
• provide the data corpus, after creating the data corpus, to cause the cognitive model to be trained based on a portion of the data corpus;
• determine a performance metric based on an output of the cognitive model;
• determine whether the performance metric satisfies a threshold corresponding to the performance metric;
• and
• selectively:
• validate the training of the cognitive model when the performance metric satisfies the threshold corresponding to the performance metric, or
• cause the cognitive model to be retrained when the performance metric does not satisfy the threshold corresponding to the performance metric.