Abstract:Sentiment polarity classification in social media is a very important task, as it enables gathering trends on particular subjects given a set of opinions. Currently, a great advance has been made by using deep learning techniques, such as word embeddings, recurrent neural networks, and encoders, such as BERT. Unfortunately, these techniques require large amounts of data, which, in some cases, is not available. In order to model this situation, challenges, such as the Spanish TASS organized by the Spanish Society for Natural Language Processing (SEPLN), have been proposed, which pose particular difficulties: First, an unwieldy balance in the training and the test set, being this latter more than eight times the size of the training set. Another difficulty is the marked unbalance in the distribution of classes, which is also different between both sets. Finally, there are four different labels, which create the need to adapt current classifications methods for multiclass handling. Traditional machine learning methods, such as Naïve Bayes, Logistic Regression, and Support Vector Machines, achieve modest performance in these conditions, but used as an ensemble it is possible to attain competitive execution. Several strategies to build classifier ensembles have been proposed; this paper proposes estimating an optimal weighting scheme using a Differential Evolution algorithm focused on dealing with particular issues that multiclass classification and unbalanced corpora pose. The ensemble with the proposed optimized weighting scheme is able to improve the classification results on the full test set of the TASS challenge (General corpus), achieving state of the art performance when compared with other works on this task, which make no use of NLP techniques.

Semi-supervised Sentiment Classification Based On Ensemble Learning with Voting

CVAE-Attention: CVAE based Semi-Supervised Sentiment Classification using Attention

Semi-Stacking for Semi-supervised Sentiment Classification.

Semi-supervised Sentiment Classification Using Ranked Opinion Words

A Mixture Language Model for the Classification of Chinese Online Reviews.

Sentiment Classification Using Supervised and Semi-Supervised Conditional Maximum Entropy Modeling.

Semi-supervised Sentiment Classification Based on Auxiliary Task Learning.

Sentiment classification: The contribution of ensemble learning

Ensemble of feature sets and classification algorithms for sentiment classification

Cooperative Hybrid Semi-Supervised Learning For Text Sentiment Classification

Semi-supervised Learning with Generative Model for Sentiment Classification of Stock Messages

Exploiting Co-occurrence Opinion Words for Semi-supervised Sentiment Classification.

Spectral Clustering-Based Semi-supervised Sentiment Classification.

Evolutionary Optimization of Ensemble Learning to Determine Sentiment Polarity in an Unbalanced Multiclass Corpus

Semi-supervised Chinese Contextual Polarity Classification with Automatic Feature Selection

A soft voting ensemble learning-based approach for multimodal sentiment analysis

Sentence-level Sentiment Classification with Weak Supervision

Data-based prediction of sentiments using heterogeneous model ensembles

Neural Sentence-Level Sentiment Classification with Heterogeneous Supervision

SPSO-EFVM: A Particle Swarm Optimization- Based Ensemble Fusion Voting Model for Sentence-Level Sentiment Analysis

Sentiment Classification in Chinese Microblogs: Lexicon-based and Learning-based Approaches