Correcting natural language processing annotators in a question answering system

摘要

An approach is provided to correct natural language processing (NLP) annotators. The approach operates by receiving a set of supporting text noted by a user in response to the user identifying an error to a user question in a question answering (QA) system. The set of supporting text includes one or more text passages from which a correct answer should have been generated by the QA system. The QA system generates one or more scored candidate corrections with each of the scored candidate corrections is based on the identified error and the set of supporting text. The user can then select one or more of the scored candidate corrections as a confirmed correction to the error. The confirmed corrections are then applied to a corpus that is utilized by the QA system when answering questions.

主权项

1. A method implemented by an information handling system that includes a memory and a processor, the method comprising:
receiving a set of supporting text noted by a user in response to the user identifying an error returned by a question answering (QA) system responsive to a question posed by the user, wherein the set of supporting text includes one or more text passages from which a correct answer should have been generated by the QA system; generating one or more candidate anomalies by comparing a context found proximate to an area of one or more of the supporting texts noted by the user to one or more similar contexts found in a corpora of the QA system; generating, by the QA system, one or more scored candidate corrections, wherein each of the scored candidate corrections is based on at least one of the generated candidate anomalies; selecting, by the user, one or more of the scored candidate corrections as a confirmed correction to the error; and applying the one or more confirmed corrections to a corpus utilized by the QA system when answering questions.