With the rapid development of the Internet, intelligent QA (Question Answering) system has been widely used in telecom operators, financial services, e-commerce shopping and other industries, but there are few researches and applications of intelligent QA system in the field of Chinese classical poetry. In view of the above situation, this paper aims to implement an automatic QA system based on the knowledge graph of Chinese classical poetry by combining natural language processing technology. In terms of the construction of knowledge graph, the common triads of Chinese classical poetry knowledge was extracted from the classical poetry websites and the knowledge graph of Chinese classical poetry stored with Neo4j was constructed. In the aspect of question recognition and multi-round dialogue, the Rasa framework was adopted to extract the entity and identify the intention of the user’s questions in Chinese classical poetry, so as to realize multi-round dialogue.
[1]
Jim Webber,et al.
A programmatic introduction to Neo4j
,
2018,
SPLASH '12.
[2]
Eric Harwit.
WeChat: social and political development of China’s dominant messaging app
,
2017
.
[3]
Justin J. Miller,et al.
Graph Database Applications and Concepts with Neo4j
,
2013
.
[4]
James W. McGuffee,et al.
Choosing Scrapy
,
2015
.
[5]
Nick Pawlowski,et al.
Rasa: Open Source Language Understanding and Dialogue Management
,
2017,
ArXiv.
[6]
Chris Eliasmith,et al.
Hyperopt-Sklearn: Automatic Hyperparameter Configuration for Scikit-Learn
,
2014,
SciPy.
[7]
Hsiu-Hsen Yao,et al.
A system for Chinese question answering
,
2003,
Proceedings IEEE/WIC International Conference on Web Intelligence (WI 2003).