Semi-Supervised Target-Dependent Sentiment Classification for Micro-Blogs

Abstract

The wealth of opinions available in the social media motivated researchers to develop automatic opinion detection tools. Many such tools are currently available online for opinion mining in short text, known as micro-blogs, but their efficacies are still limited. Current tools focus on detecting sentiment polarity expressed in a micro-blog regardless of the topic (target) discussed. Little improved approaches have been proposed to detect sentiment towards a specific target, referred to as target-dependent sentiment classification. Our literature review has shown that all these target-dependent approaches use supervised learning techniques. Such techniques need a huge amount of labeled data for increasing classification accuracy. However, preparing labeled data from social media needs a lot of efforts. In this work, we address this issue by employing semisupervised learning techniques that have not been used before with target-dependent sentiment classification. To the best of our knowledge, our work is the first research that employs semisupervised learning techniques in this direction. Semi-supervised learning techniques have been known in the literature to improve classification accuracy in comparison with supervised learning techniques; however, they use same number of labeled samples plus many unlabelled ones. In this work, we propose a new semi-supervised learning technique that uses less number of labeled microblogs than that used with supervised learning techniques. Experiment results have shown that the proposed technique provides competitive accuracy.

Hybrid Approach for User Reviews' Text Analysis and Visualization: A Case Study of Amazon User Reviews

“…The second type is generated from reviews that have 4-star ratings are labeled as positive and negative ones are generated from 1-star and 2-star reviews. In study of [15] the authors proposes semi-supervised learning targeted sentiment classification by using both labelled and unlabeled data. The analysis of text [16] on the web can also be unsupervised where unorganized groups of data, and by using artificial intelligence, the models are trained in order to understand them in a comprehensive way.…”

Creación de un Corpus de Opiniones con Emociones Usando Aprendizaje Automático

“…El hecho que estos algoritmos semi supervisados utilicen pocas instancias etiquetadas para entrenar hace que que sean atractivos de utilizar en aquellos casos donde el etiquetado sea muy costoso, requiera tiempo o intervención humana. En (Abudalfa and Ahmed, 2019) se evaluaron una serie de algoritmos semi supervisados aplicados a textos de micro-blogs y en idioma inglés, concluyendo que este enfoque es relevante para la clasificación textual.…”

Identificación de emociones en textos de una red social

“…Los algoritmos semisupervisados son atractivos para utilizar en aquellos casos donde el etiquetado es muy costoso y requiera tiempo o intervención humana. En Abudalfa & Ahmed (2019), se evaluaron una serie de algoritmos semisupervisados aplicados a textos de micro-blogs y en idioma inglés, y se concluyó que este enfoque es relevante para la clasificación textual.…”