基于深度学习的藏文舆情分析研究

Research on Tibetan Public Opinion Analysis Based on Deep Learning

随着自然语言处理技术的快速发展,藏文信息处理技术也取得了较大进展。其中,藏文舆情分析作为藏族地区社会舆情分析的重要技术,受到广泛关注。但是,现有的藏文文本情感分析研究由于起步较晚,还存在很大提升空间。本文提出基于深度集成学习的藏文文本情感分析算法。并通过建立藏文情感数据集进行实验,本文算法在三类情感(正向、负向、中性)中精确率平均提升1.65%,召回率提升1.63%,F1分数提升1.96%。实验结果表明,本文采用的深度集成学习有效地提高了文本情感分类的性能。

With the rapid development of natural language processing technology, Tibetan information processing technology has also made great progress. Among them, Tibetan public opinion analysis, as an important technology for social public opinion analysis in Tibetan areas, has received widespread attention. However, due to the late start of the existing Tibetan text sentiment analysis research, there is still a lot of room for improvement. This paper proposes a Tibetan text sentiment analysis algorithm based on deep ensemble learning. Moreover, through the establishment of Tibetan emotion datasets for experiments, the algorithm in this paper has an average increase of 1.65% in precision, 1.63% in recall, and 1.96% in F1 scores in the three types of emotions(positive, negative, and neutral).