Emotion Detection from text: A Survey

Abstract

This survey describes recent works in the field of Emotion Detection from text, being a part of the broader area of Affective Computing. This survey has been inspired on the well-known fact that, despite there is a lot of work on emotional detection systems, a lot of work is expected to be done yet. The increment of these systems is due to the large amount of emotional data available in Social Web. Detecting emotions from text have attracted the attention of many researchers in computational linguistics because it has a wide range of applications, such as suicide prevention or measuring well-being of a community. This paper mainly collects works based on lexical and machine learning approaches and these works are classificated in accordance with the emotional model and the approach used.

Emotion and sentiment analysis from Twitter text

“…According to the previous surveys [37,51,54,52,53] tried natural language processing [58], linguistic rule-based methods [76], ensemble of multiple methods [13] or some novel methods almost completely unique and achieved some good results. Modifications to those novel methods may lead to an improved and highly accurate automatic emotion detection system from text input.…”

“…On the other hand, Dimensional emotion models define a few dimensions with some parameters and specify emotions according to those dimensions. Two or three dimensions are used in most dimensional emotion models -'valence' (indicates the positivity or negativity of an emotion), 'arousal' (indicates the excitement level of an emotion) and 'dominance' (indicates the level of control over an emotion) [36,37].…”

KGPChamps at SemEval-2019 Task 3: A deep learning approach to detect emotions in the dialog utterances.

“…A detail description of different hidden challenges present in emotion detection over social media content is present in (Dini and Bittar, 2016). Few survey papers (Canales and Martínez-Barco, 2014;Seyeditabari et al, 2018) describing different emotion analysis and detection methods adopted in past years also came up during this period.…”

CLaC Lab at SemEval-2019 Task 3: Contextual Emotion Detection Using a Combination of Neural Networks and SVM

“…With the advent of social media, emotion detection from text has been used to track bloggers' mental health and has been explored using different techniques, such as lexicon-based approaches and machine learning (Canales and Martínez-Barco, 2014). Lexicon-based approaches include keyword-based and ontological approaches.…”