All-in-One: Emotion, Sentiment and Intensity Prediction Using a Multi-Task Ensemble Framework

We propose a multi-task ensemble framework that jointly learns multiple related problems. The ensemble model aims to leverage the learned representations of three deep learning models (i.e., CNN, LSTM and GRU) and a hand-crafted feature representation for the predictions. Through multi-task framewor...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on affective computing 2022-01, Vol.13 (1), p.285-297
Main Authors: Akhtar, Md Shad, Ghosal, Deepanway, Ekbal, Asif, Bhattacharyya, Pushpak, Kurohashi, Sadao
Format: Article
Language:English
Subjects:
Arousal
arousal prediction
Classification
coarse-grained emotion analysis
Computational modeling
Data mining
Deep learning
dominance prediction
Emotion analysis
Emotions
ensemble
fine-grained emotion analysis
fine-grained sentiment analysis
Hidden Markov models
intensity prediction
Machine learning
multi-layer perceptron
Organizations
Predictive models
Representations
Sentiment analysis
Task analysis
valence prediction
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We propose a multi-task ensemble framework that jointly learns multiple related problems. The ensemble model aims to leverage the learned representations of three deep learning models (i.e., CNN, LSTM and GRU) and a hand-crafted feature representation for the predictions. Through multi-task framework, we address four problems of emotion and sentiment analysis, i.e., "emotion classification & intensity ", " valence , arousal & dominance for emotion", " valence & arousal for sentiment", and " 3-class categorical & 5-class ordinal classification for sentiment". The underlying problems cover two granularity (i.e., coarse-grained and fine-grained ) and a diverse range of domains (i.e., tweets , Facebook posts , news headlines , blogs , letters etc.). Experimental results suggest that the proposed multi-task framework outperforms the single-task frameworks in all experiments.
ISSN:1949-3045
1949-3045
DOI:10.1109/TAFFC.2019.2926724