期刊名称:International Journal of Advanced Computer Science and Applications(IJACSA)
印刷版ISSN:2158-107X
电子版ISSN:2156-5570
出版年度:2019
卷号:10
期号:9
页码:60-68
出版社:Science and Information Society (SAI)
摘要:Customer support has become one of the most
important communication tools used by companies to provide
before and after-sale services to customers. This includes
communicating through websites, phones, and social media
platforms such as Twitter. The connection becomes much faster
and easier with the support of today's technologies. In the field of
customer service, companies use virtual agents (Chatbot) to
provide customer assistance through desktop interfaces. In this
research, the main focus will be on the automatic generation of
conversation “Chat” between a computer and a human by
developing an interactive artificial intelligent agent through the
use of natural language processing and deep learning techniques
such as Long Short-Term Memory, Gated Recurrent Units and
Convolution Neural Network to predict a suitable and automatic
response to customers’ queries. Based on the nature of this
project, we need to apply sequence-to-sequence learning, which
means mapping a sequence of words representing the query to
another sequence of words representing the response. Moreover,
computational techniques for learning, understanding, and
producing human language content are needed. In order to
achieve this goal, this paper discusses efforts towards data
preparation. Then, explain the model design, generate responses,
and apply evaluation metrics such as Bilingual Evaluation
Understudy and cosine similarity. The experimental results on
the three models are very promising, especially with Long ShortTerm
Memory and Gated Recurrent Units. They are useful in
responses to emotional queries and can provide general,
meaningful responses suitable for customer query. LSTM has
been chosen to be the final model because it gets the best results
in all evaluation metrics.
关键词:Chatbot; deep learning; natural language
processing; similarity