Text Mining Analysis on Users' Reviews for News Aggregator Toutiao




Li Jia, Wang Bifeng, Ni Alexandra Jingsi, Liu Qian

Laura Tan, Jianglian Zhao

International Conference on Artificial Intelligence and Cloud Computing

PublisherIOP Publishing Ltd

2021

Journal of Physics: Conference Series

International Conference on Artificial Intelligence and Cloud Computing (ICAICC) 2020 18-20 December 2020, Suzhou, China

Journal of Physics: Conference Series

Journal of Physics: Conference Series

1771

012008

1742-6588

DOIhttps://doi.org/10.1088/1742-6596/1771/1/012008(external)

https://iopscience.iop.org/article/10.1088/1742-6596/1771/1/012008(external)

https://research.utu.fi/converis/portal/detail/Publication/54797029(external)



This paper intent to improve consumer communication by text mining analysis
with users’ reviews. The news aggregator we focus on is: Toutiao, known as “today's
headlines” in Chinese. It is the top news aggregator application run by Bytedance
company in China. It utilizes AI algorithms to provide numerous news feed for its users. As new technologies are shaping the business strategy studies as well as online
communication analysis, it requires innovative and effective analyses of unconventional
data, such as the 12,290 online reviews on Toutiao we collected from Apple’s App Store. Through the LDA topic modelling and sentiment analysis, our research has identified three
major negative complains the consumers have regarding Toutiao application, namely: too
many unsolicited advertisements, contents (vulgar content, time consuming video, privacy
and copyrights infringement issue) and incompatibility with the latest Apple digital
devices


Last updated on 2024-26-11 at 15:55