A hybrid model for aspect-based sentiment analysis on customer feedback: research on the mobile commerce sector in Vietnam

Feedback and comments on mobile commerce applications are extremely useful valuable information sources that reflect the quality of products or services to determine whether data is positive negative help businesses monitor brand product sentiment in customers’ feedback understand needs. However, increasing number makes it increasingly difficult customers using manual methods. To solve this problem, study builds a hybrid research model based aspect mining comment classification for aspect-based analysis (ABSA) deeply comprehend customer their experiences. Based previous results, we first construct dictionary words e-commerce field. Then, POS tagging technique applied word Vietnamese extract aspects related words. The implemented with machine deep learning methods corpus comprising more than 1,000,000 opinions collected from Vietnam's four largest applications. Experimental results show Bi-LSTM method has highest accuracy 92.01%; selected proposed analyze viewpoint real data. findings can be online experience time, enable administrators make timely accurate decisions, improve take competitive advantage.