POSpart of speech METHODS FOR PERFORMING SENTIMENT ANALYSIS OF MESSAGES IN SOCIAL NETWORK SERVICE BASED ON PART OF SPEECH FEATURE AND SENTIMENT ANALYSIS APPARATUS FOR PERFORMING THE SAME

摘要

A POS feature-based emotion analysis method performed in an emotion analysis device for emotion analysis of a social network service (SNS) message includes the steps of selecting at least one sentence of sentences in an SNS message as a training data set, and the selected training data Among the sentences in the set, sentences that are low or not related to the emotional feature are removed from the training dataset, and words in the sentences included in the training dataset are POS tagging rules. Grouping into a feature set by tagging according to, and assigning weights to words in sentences included in the selected training data set based on the dependency of the clustered feature set, and the Grouping sentences included in the training dataset into classes having at least two polarities, positive and negative, wherein the sentences included in the training dataset are classified into positive sentences and negative sentences according to the polarity. do.