Abstract: A method for computer question answering includes, at a retriever subsystem of a question answering computer system, identifying a plurality of relevant text evidence strings for an input text question. At a linker subsystem of the question answering computer system, one or more of the plurality of relevant text evidence strings are associated with a respective secondary text evidence string to form a plurality of evidence chains via a previously-trained entity-linking machine-learning model. At a chainer subsystem of the question answering computer system, a ranked set of the evidence chains is identified based at least in part on an output of a generative machine-learning model applied to each of the plurality of evidence chains. At a reader subsystem of the question answering computer system, an answer to the input text question is output based at least in part on the ranked set of evidence chains.

Abstract: A computer-implemented system and method relates to natural language processing. The computer-implemented system and method are configured to obtain a current data structure from a global knowledge graph, which comprises various knowledge graphs. The current data structure includes a current head element, a current relationship element, and a current tail element. A sentence is obtained based on the current data structure. A question is generated by removing the current tail element from the sentence. A correct answer is generated for the question. The correct answer includes the current tail element. A pool of data structures is extracted from the global knowledge graph based on a set of distractor criteria. The set of distractor criteria ensures that each extracted data structure includes the current relationship element. Tail elements from the pool of data structures are extracted to create a pool of distractor candidates. A set of distractors are selected from the pool of distractor candidates.