The traditional text sentiment analysis directly inputs syntactic feature or word vector to model. It fails to consider the characteristics of time series in sentiment. Considering that product reviews are short text, this paper comes up with the method of windowed word vector and classifier fusion in the decision layer. The results indicate that the proposed method can achieve better performance than several existing methods.
CITATION STYLE
Zhao, D., Shen, Y., Shen, Y., Ma, Y., Jin, Y., Li, S., & Gu, M. (2020). Short-text sentiment analysis based on windowed word vector. In Lecture Notes in Electrical Engineering (Vol. 517, pp. 168–174). Springer Verlag. https://doi.org/10.1007/978-981-13-6508-9_22
Mendeley helps you to discover research relevant for your work.