An Overseas Sentiment Analysis of Xi’an City Based on Deep Learning and NOW Corpus

With the development of corpus linguistic and the inter-disciplinary application of deep learning, the quantitative research on the cross-cultural communication of city image is becoming more and more popular recently. The purpose of this study is to reveal the current communicative situation of Xi’an City on English online news medias and overseas audiences. This paper uses the real-time English online news Corpus NOW and public opinion texts drawn from Twitter as the research data. Firstly, this study analyzes the text data from the perspective of Word Frequency, Keywords, Key Word In Context (KWIC) and Collocation. Then sentiment analysis based on a pre-trained Natural Language Processing (NLP) model Bidirectional Encoder Representations from Transformers (BERT) is used to analyze overseas audiences’ attitude toward Xi’an City. The results find that Xi’an City is less reported by overseas online new medias than other 14 new first-tier cities in China. Southeast and Central Asian countries are the main reporters of Xi’an City. International society mainly concentrates on Xi’an City’s national defense industry (especially military aircraft), followed by its long history and economy. Besides, 46.2% of overseas audiences keeps neutral toward Xi’an, 30.4% have a negative attitude, and another 23.4% have a positive attitude. According to the research result, this study proposes three practical suggestions to help enhance Xi’an City’s international influence effectively.