Open AccessResearch and Implementation of Seq2Seq Model Chat Robot Based on Attention Mechanism
Author(s) -
Weiyao Luo,
Di Wu
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1693/1/012200
Subject(s) - computer science , mechanism (biology) , word2vec , artificial intelligence , natural language processing , robot , word (group theory) , human–computer interaction , linguistics , embedding , philosophy , epistemology
With the continuous development of deep learning technology, computers can not only understand the natural language input by users, but also reply to the sentences used for input. This paper designs an intelligent chat robot based on Attention mechanism and Seq2Seq model. It uses jieba word segmentation tool and Word2vec to convert sentences in corpus into semantic vectors, and then trains and tests the model through TensorFlow platform. The experiment compares whether the Attention mechanism is added or not, and proves that the model added with Attention mechanism has better effect than the traditional Seq2Seq model, and performs better in actual dialogue.