Capturing Global Mood Levels using Blog Posts

This paper builds models that predict the levels of various moods according to the language used by bloggers at a given time; these models show high correlation with the moods actually measured, and substantially outperform a baseline.

Abstract

The personal, diary-like nature of blogs prompts many bloggers to indicate their mood at the time of posting. Aggregating these indications over a large amount of bloggers gives a “blogosphere state-of-mind ” for each point in time: the intensity of different moods among bloggers at that time. In this paper, we address the task of estimating this state-of-mind from the text written by bloggers. To this end, we build models that predict the levels of various moods according to the language used by bloggers at a given time; our models show high correlation with the moods actually measured, and substantially outperform a baseline.