Question Answering with Deep Learning: A Survey
Date Issued
2019-05
Author(s)
Toshevska, Martina
Mirceva, Georgina
Abstract
Automatically generating answer for a given question is a process in which the computer is supposed to answer a
question in a natural language where the question itself is also
provided in natural language. Deep learning techniques gained
extensive research in both fields of computer vision and natural
language processing. Therefore, they are extensively applied for
the task of question answering using wide varieties of datasets.
This survey aims to overview some of the latest algorithms
and models proposed in the field, as well as datasets exploited
for training and evaluating the models. In this survey, the models
are presented as part of one of the following groups: classical
deep neural networks, dynamic memory networks and relation
networks. Several datasets have been proposed specifically for
the research on automatic question answering. This survey
briefly overviews datasets for two different categories of question
answering: textual and visual. In the end, evaluation metrics
utilized in the field are presented, grouped as: metrics for
evaluation of an information retrieval system and metrics for
evaluating automatically generated text.
question in a natural language where the question itself is also
provided in natural language. Deep learning techniques gained
extensive research in both fields of computer vision and natural
language processing. Therefore, they are extensively applied for
the task of question answering using wide varieties of datasets.
This survey aims to overview some of the latest algorithms
and models proposed in the field, as well as datasets exploited
for training and evaluating the models. In this survey, the models
are presented as part of one of the following groups: classical
deep neural networks, dynamic memory networks and relation
networks. Several datasets have been proposed specifically for
the research on automatic question answering. This survey
briefly overviews datasets for two different categories of question
answering: textual and visual. In the end, evaluation metrics
utilized in the field are presented, grouped as: metrics for
evaluation of an information retrieval system and metrics for
evaluating automatically generated text.
Subjects
File(s)![Thumbnail Image]()
Loading...
Name
QuestionAnsweringwithDeepLearning-ASurvey.pdf
Size
134.97 KB
Format
Adobe PDF
Checksum
(MD5):15316996260e014517a93f73bfef8c39