Dataset for modeling Beck’s cognitive triad to understand depression

Abstract

This article presents data to model Beck’s cognitive triad to understand the subjective symptoms of depression, such as negative view of self, future, and world. The Cognitive Triad Dataset (CTD) comprises 5886 messages, 600 from the Time-to-Change blog, 580 from Beyond Blue personal stories, and 4706 from Twitter. The data were manually labeled by skilled annotators. This data is divided into six categories: self-positive, world-positive, future-positive, self-negative, world-negative, and future-negative. The Cognitive Triad Dataset was evaluated on two subtasks: aspect detection and sentiment classification on given aspects. The dataset will aid in the comprehension of Beck’s Cognitive Triad Inventory (CTI) items in a person’s social media posts.

Keywords

• Cognitive triad
• Depression
• Sentiment classification