Sentiment Analysis with Improved Adaboost and Transfer Learning Based on Gaussian Process

Sentiment analysis is an increasingly important area in NLP to extract opinions and sentiment expressed by humans. Traditional methods are often difficult to tackle the problems of different sample distribution and domain dependence, which seriously limits the development of sentiment classification. In this paper, a novel sentiment analysis method is proposed by combining improved Adaboost and transfer learning based on Gaussian Processes to solve these two problems. A Paragraph Vector Model is employed to obtain the continuous distributed vector representations. Then, Adaboost method is used to choose the most important training features in source training data and auxiliary data. Finally, an asymmetric transfer learning classifier is introduced in Gaussian Processes. It is shown that, compared with the existing algorithms, our method is more effective for the different sample distribution and domain dependence.