Search

In recent years, Aspect-based Sentiment Analysis (ABSA) has been a crucial yet challenging task in recognizing textual emotions from text. ABSA has numerous application across various fields, such as social media, commodity review, and movie comment, making it an attractive area of research. Many researchers are working to develop more powerful sentiment analysis models. Currently, most existing ABSA models use the generic pre-trained language models (PLMs) based fine-tuning paradigm, which only utilizes the encoder parameters while discarding the decoder parameters of PLMs. However, this approach fails to leverage the prior knowledge revealed in PLMs effectively. To address these issues, we investigate the potential of the initial pre-training scheme of PLMs to conduct ABSA and thus propose a novel approach in this paper, namely Target Word Transferred ABSA (WordTransABSA). In WordTransABSA, we propose "Word Transferred LM", a novel sequence-level optimization strategy that transferred target words in sentence into pivot tokens to stimulate better PLM semantic understanding capability. Given a sentence with aspect terms as input, WordTransABSA generates contextually appropriate semantics and predicts the affective tokens on the corresponding positions of the aspect terms. The final sentiment polarity of each aspect term is determined through several sentiment identification strategies that we selected. WordTransABSA takes full advantage of the versatile linguistic knowledge of Pre-trained Language Model, resulting in competitive accuracy compared with recent baselines. The WordTransABSA demonstrates its superiority and effectiveness through extensive experiments in both data-sufficient (full-data supervised learning) and data-insufficient (few-shot learning) scenarios. We have made our code publicly available on GitHub: https://github.com/albert-jin/WordTransABSA. • We present WordTransABSA, introducing a "Word Transferred LM" strategy for ABSA. • Various strategies are explored for searching sentiment-related pivot tokens. • Data-sufficient/scarce experiments validate the effectiveness of WordTransABSA. • Stimulating PLM pre-training paradigm is proved to be a better solution for ABSA. [ABSTRACT FROM AUTHOR]

The usage of digital/social media is increasing day by day with the advancement of technology. People in the twenty-first century are being raised in an internet-enabled world with social media. Communication has been just one button click. Even though there are plenty of opportunities with digital media people tend to misuse it. People spread hatred toward a person in social networking. Cyberbullying affects people in different aspects. It doesn't affect only for health, there are more different aspects which will lead life to a threat. Cyberbullying is a worldwide modern phenomenon which humans cannot avoid hundred percent but can be prevented. Most existing solutions have shown techniques/approaches to detect cyberbullying, but they are not freely available for end-users to use. They haven't considered the evolution of language which makes a big impact on cyberbullying text. This paper describes a system for automatic detection and prevention cyberbullying considering the main characteristics of cyberbullying such as Intention to harm an individual, Repeatedly and over time and using abusive curl language or hate speech using supervised machine learning. The system relies on the detection of cyberbullying text along with the themes/categories associated with cyberbullying such as racist, sexual, physical mean, swear and other, using support vector machines and Logistic regression. The author of this research presents a new hypothesis for cyberbullying detection that the circumstances and usage of texting and its language have changed by time. Most of the studies have considered calling someone stupid, ugly and idiot as cyberbullying. Things have changed, such words may or may not always be a bullying incident. If a person wants intentionally to harm an individual, they will use extreme words. In addition to traditional feature extraction techniques like Term Frequency–Inverse Document Frequency (TF-IDF), N-gram and profanity along with sentiment analysis increases the accuracy of the system. Evaluated the proposed system using Recall, Precision and F1-score. [ABSTRACT FROM AUTHOR]

This paper presents a supervised learning method for irony detection in Arabic tweets. A binary classifier uses four groups of features whose efficiency has been empirically proved in other languages such as French, English, Italian, Dutch and Japanese. Our first results are encouraging and show that state of the art features can be successfully applied to Arabic language with an accuracy of 72.76%. [ABSTRACT FROM AUTHOR]

• A fuzzy rule-based approach is proposed for sentiment analysis of tweets. • Formulation of nine fuzzy rules to compute sentiment of each tweet. • The proposed unsupervised approach is suitable for any sentiment lexicon. • The approach can be adapted to both bipolar and tripolar sentiment analysis tasks. In this paper, we compute the sentiment of social media posts using a novel set of fuzzy rules involving multiple lexicons and datasets. The proposed fuzzy system integrates Natural Language Processing techniques and Word Sense Disambiguation using a novel unsupervised nine fuzzy rule based system to classify the post into: positive, negative or neutral sentiment class. We perform a comparative analysis of our method on nine public twitter datasets, three sentiment lexicons, four state-of-the-art approaches for unsupervised Sentiment Analysis and one state-of-the-art method for supervised machine learning. Traditionally, Sentiment Analysis of twitter data is performed using a single lexicon. Our results can give an insight to researchers to choose which lexicon is best for social media. The fusion of fuzzy logic with lexicons for sentiment classification provides a new paradigm in Sentiment Analysis. Our method can be adapted to any lexicon and any dataset (two-class or three-class sentiment). The experiments on benchmark datasets yield higher performance for our approach as compared to the state-of-the-art. [ABSTRACT FROM AUTHOR]