Text Segmentation based on One Hot and Word Vector Representation

EasyChair Preprint no. 1926

Abstract

The sentiment orientation analysis aims to find out the public's attitude towards something and is widely used in product analysis and public opinion detection. This paper uses real text corpus to achieve automatic judgment of sentiment orientation through preprocessing, text coding representation, feature extraction, classification, etc.; explores the unique heat expression and word vector representation of text, neighbor, naive Bayes, support vector machine The effect of the algorithm on the judgment of sentiment orientation. Through multiple sets of comparative experiments, we found that the word vector representation of the text significantly improved the accuracy and speed of the sentiment analysis. At the same time, for the case of less training corpus, we try to carry out the word vector model "secondary training", and the experiment shows that the word vector quality can be effectively improved. In summary, the process of this paper has a certain effect on the analysis of text sentiment orientation.

Keyphrases: one hot, text sentiment orientation, Word Vector Representation

@Booklet{EasyChair:1926,
  author = {Ruoqi Dang},
  title = {Text Segmentation based on One Hot and Word Vector Representation},
  howpublished = {EasyChair Preprint no. 1926},

  year = {EasyChair, 2019}}