A novel LMAEB-CNN model for Chinese microblog sentiment analysis
- PDF / 1,527,117 Bytes
- 15 Pages / 439.37 x 666.142 pts Page_size
- 12 Downloads / 193 Views
A novel LMAEB‑CNN model for Chinese microblog sentiment analysis Yi‑Jen Su1 · Wu‑Chih Hu2 · Ji‑Han Jiang3 · Ruei‑Ye Su1
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Recent emergence and flourishment of social media services (SMS) have led to drastic changes in the way people interact. When Internet users share their feelings and opinions about some specific topic online, the massive amount of discussion forms a real-time public opinion pool. Given this phenomenon, employing sentiment analysis to measure public opinions from online social communities has become a hot research issue in these years. To achieve higher classification accuracy in Chinese sentiment analysis, this research proposes the LMAEB-CNN model, which combines Bi-LSTM and CNN with multi-head attention mechanism. The proposed model not only solves the over-fitting problem, but also promotes the accuracy of emotional polarity classification. To carry out experiments, we used datasets collected from four famous Chinese SMS platforms, including Plurk, PTT, Dcard, and Mobile01, and the LMAEB-CNN model demonstrated higher accuracy than other methods, including SVM, CNN, LSTM, and AEB-CNN. Keywords Sentiment analysis · Attention mechanism · Long short-term memory · Convolutional neural network
* Yi‑Jen Su [email protected] Wu‑Chih Hu [email protected] Ji‑Han Jiang [email protected] Ruei‑Ye Su [email protected] 1
Computer Science and Information Engineering, Shu-Te University, Yanchao Dist., Kaohsiung City, Taiwan
2
Computer Science and Information Engineering, National Penghu University of Science and Technology, Magong City, Penghu County, Taiwan
3
Computer Science and Information Engineering, National Formosa University, Huwei Township, Yunlin County, Taiwan
13
Vol.:(0123456789)
Y.-J. Su et al.
1 Introduction With these years network usage becoming popular, network awareness has become an important index in predicting the future trend or development of an event. It is the hope of many government organizations, political parties and business companies to collect real-time opinions of the common people and be able to make subsequent adjustments to their new policies or new commercial products. When conducting public opinion measurement, the time needed for collecting public opinion data can effectively be shortened by collecting published short messages on SMS platforms to evaluate the network users’ opinion (classified as positive, negative, or neutral) on a specific event or news. After the data gathering process, the subsequent sentiment analysis [1] result can be used to predict the future outcome of a specific event, such as election results, stock price fluctuation, product sales effectiveness, and so on. A high-precision Chinese microblog sentiment analysis can serve many kinds of functions. Generally, the public opinion monitoring process includes massive amounts of microblogs collection, data cleaning and preprocessing, and sentiment analysis. Based on the length of the to-process text, the se
Data Loading...
