AUTHOR’S OPINION MINING USING INVERSE DOCUMENT FREQUENCY

Abstract

The purpose of the article is to identify patterns for author’s opinion mining in articles relating to the genre of information journalism, that imposes a restriction on the expression of judgments and opinions. A corpus of political articles has been assembled by the author and the metric from information retrieval called inverse document frequency has been counted for every token in the collection. The applicability of this criterion has been shown for the search of colloquial language which is an indicator of the presence of the author’s opinion. In conclusion, the directions of further research of this problem are formulated and the question of identifying the author’s opinion through the inverse document frequency in other texts is considered.

Keywords

• sentiment analysis
• information retrieval
• inverse document frequency
• term frequency
• opinion