Method for training machine reading comprehension model, computer-readable recording medium and question answering system

A computer-readable recording medium storing a program for implementing a method for training a machine reading comprehension model, the method comprising:

• generating an input sequence including a plurality of tokens based on at least one feature among features of input sentences;
• embedding the input sequence using a pre-trained language model;
• inputting an output embedding, obtained by embedding the input sequence, to a first encoding layer to perform word-level encoding;
• inputting a portion of hidden states of the first encoding layer to a second encoding layer to perform sentence-level encoding;
• calculating a first loss value for a starting position of the answer span and a second loss value for an ending position of the answer span;
• and
• calculating a third loss value for a position of a sentence including the answer span.

A question answering system, comprising:

• a communication module configured to receive a question sentence of a user;
• and
• a question answering module configured to search for an answer span to the question sentence using a trained machine reading comprehension model;
• wherein the trained machine reading comprehension model comprises:
• an embedding layer configured to embed an input sequence corresponding to the question sentence using a pre-trained language model;
• a first encoding layer configured to perform word-level encoding on an output embedding obtained by embedding the input sequence;
• a second encoding layer configured to perform sentence-level encoding on a portion of hidden states of the first encoding layer;
• and
• a loss value calculator configured to calculate a first loss value for a starting position of the answer span, a second loss value for an ending position of the answer span, and a third loss value for a position of a sentence including the answer span.