Aspect-based Sentiment Analysis (ABSA) using Machine Learning Algorithms DOI
Ayesha Siddiqua,

V Bindumathi,

Ganesh Raghu

et al.

Published: April 26, 2024

Aspect-Based Sentiment Analysis (ABSA) is a Natural Language Processing task that aims to identify and extract the sentiment of specific aspects or components product service. ABSA typically involves multi-step process begins with identifying features service are being discussed in text. This followed by analysis, where polarity (positive, negative, neutral) assigned each aspect based on context sentence document. Finally, results aggregated provide an overall for aspect. The training machine learning models classify text neutral). First, we transform data using Term Frequency-Inverse Document Frequency (TF-IDF), which assigns weights words their importance within document collection. emphasizes informative terms. Then, these TF-IDF fed into both SVM Logistic Regression models. find hyper plane best separates classes, while calculates probability belonging class. Extensive experiments have been conducted datasets covid vaccinations dataset show support vector model achieves excellent performance terms extraction classification. Twitter can be imbalanced, more positive negative tweets depending topic. affect process. Techniques like oversampling undersampling minority class might necessary. work investigates algorithms classification task. Support Vector Machine (SVM) (LR) were compared. indicate achieved superior accuracy (87.34%) compared (84.64%), suggesting as suitable option this

Language: Английский

Agenda-setting effects for covid-19 vaccination: Insights from 10 million textual data from social media and news articles using BERTopic DOI
Hyunsang Son, Young Eun Park

International Journal of Information Management, Journal Year: 2025, Volume and Issue: 83, P. 102907 - 102907

Published: April 8, 2025

Language: Английский

Citations

0
Recent development on online public opinion communication and early warning technologies: Survey DOI
Wei Wu, Yawen Yang, Tianlu Qiao

et al.

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127823 - 127823

Published: April 1, 2025

Language: Английский

Citations

0
Co-Bi-LSTM: Opinion Mining on Twitter Data Using Convolutional Neural Network with Optimized Bidirectional LSTM Model DOI Open Access

Bikshapathy Peruka,

K. Shahu Chatrapati

International Journal of Computational and Experimental Science and Engineering, Journal Year: 2025, Volume and Issue: 11(2)

Published: May 13, 2025

The proliferation of social networking platforms has generated a substantial volume user-generated content, posing significant challenges for text classification due to its diverse nature. Sentiment analysis or opinion mining, is crucial extracting insights from user opinions and emotions regarding various entities events. This research classifies tweets into positive negative sentiments using Twitter data predictive in domains such as consumer behaviour election outcomes. Two Kaggle datasets like Sentiment140 News Headlines Dataset Sarcasm Detection are used. pre-processing phase includes cleaning, tokenization, padding. Word embedding skip-gram can capture semantic relationships used neural network architectures word2vec conversion. paper proposes hybrid model called Convolutional Optimized Bidirectional LSTM (CO-Bi-LSTM), combining Neural Networks (CNN) with an Long Short-Term Memory (O-Bi-LSTM) network, enhanced by the Hybrid Hippopotamus based Zebra Optimization Algorithm (HH-ZOA). model's performance evaluated metrics accuracy, F-measure, precision, demonstrating efficacy sentiment data.

Language: Английский

Citations

0
A review on the use of machine learning techniques in monkeypox disease prediction DOI Creative Commons
Shailima Rampogu

Science in One Health, Journal Year: 2023, Volume and Issue: 2, P. 100040 - 100040

Published: Jan. 1, 2023

Infectious diseases have been posing to be a global threat in the recent times and are progressing from endemics pandemics. The early detection finding better cure is one method curb disease its transmission. advent of machine learning (ML) demonstrate ideal approach diagnosis disease. In current review, use ML algorithm monkeypox (MP) highlighted. To extract useful information dataset, various models like CNN, DL, NLP, Naive Bayes, GRA-TLA, HMD, ARIMA, SEL, Regression analysis, Twitter postings were built. These findings show that detection/classification, forecast sentiment analysis primarily analyzed. Furthermore this review will assist researchers understanding latest implementation on MP further progress field discover potent therapeutics.

Language: Английский

Citations

8
Explainable assessment of financial experts’ credibility by classifying social media forecasts and checking the predictions with actual market data DOI Creative Commons
Silvia García-Méndez, Francisco de Arriba-Pérez, Jaime González‐González

et al.

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 255, P. 124515 - 124515

Published: June 15, 2024

Social media include diverse interaction metrics related to user popularity, the most evident example being number of followers. The latter has raised concerns about credibility posts by popular creators. However, existing approaches assess in social strictly consider this problem a binary classification, often based on priori information, without checking if actual real-world facts back users' comments. In addition, they do not provide automatic explanations their predictions foster trustworthiness. work, we propose assessment solution for financial creators that combines Natural Language Processing and Machine Learning. reputation contributors is assessed automatically classifying forecasts asset values type verifying these with market data approximate probability success. outcome verification continuous score instead result, an entirely novel contribution work. Moreover, (i.e., context) are exploited calculating correlation rankings, providing insights interest end-users drop or rise). Finally, system provides natural language its decisions model-agnostic analysis relevant features.

Language: Английский

Citations

2
A Pre-Trained Model for Aspect-based Sentiment Analysis Task: using Online Social Networking DOI Open Access
Amit Chauhan, Aman Sharma, Rajni Mohana

et al.

Procedia Computer Science, Journal Year: 2024, Volume and Issue: 233, P. 35 - 44

Published: Jan. 1, 2024

Aspect-based sentiment analysis (ABSA) is a crucial part of Natural Language Processing (NLP) that focuses on identifying emotions related to specific elements in written material. ABSA has gained widespread interest due its ability provide precise insights into expressions across different domains. Social media provides valuable resource for ABSA, containing user-created content with viewpoints and feedback. However, the informal nature social text poses challenges ABSA. This study investigates performance enhancement baseline proposed models task context. Both were evaluated accuracy F1 score improvements. The results showed suggested model performs better than other models, an improvement 0.52% overall increase 1.16%. analyses laptops indicated limitations model's performance, scores ranging from 72.65% 84.98%.

Language: Английский

Citations

2
UrduAspectNet: Fusing Transformers and Dual GCN for Urdu Aspect-Based Sentiment Detection DOI Open Access
Kamran Aziz, Aizihaierjiang Yusufu, Jun Zhou

et al.

ACM Transactions on Asian and Low-Resource Language Information Processing, Journal Year: 2024, Volume and Issue: unknown

Published: May 4, 2024

Urdu, characterized by its intricate morphological structure and linguistic nuances, presents distinct challenges in computational sentiment analysis. Addressing these, we introduce ”UrduAspectNet” – a dedicated model tailored for Aspect-Based Sentiment Analysis (ABSA) Urdu. Central to our approach is rigorous preprocessing phase. Leveraging the Stanza library, extract Part-of-Speech (POS) tags lemmas, ensuring Urdu’s intricacies are aptly represented. To probe effectiveness of different embeddings, trained using both mBERT XLM-R comparing their performances identify most effective representation Urdu ABSA. Recognizing nuanced inter-relationships between words, especially flexible syntactic constructs, incorporates dual Graph Convolutional Network (GCN) layer.Addressing challenge absence ABSA dataset, curated own, collecting over 4,603 news headlines from various domains, such as politics, entertainment, business, sports. These headlines, sourced diverse platforms, not only prevalent aspects but also pinpoints polarities, categorized positive, negative, or neutral. Despite inherent complexities colloquial expressions idioms, showcases remarkable efficacy. Initial comparisons embeddings integrated with GCN provide valuable insights into respective strengths context With broad applications spanning media analytics, business insights, socio-cultural analysis, positioned pivotal benchmark research.

Language: Английский

Citations

2
Analyzing Public Reactions, Perceptions, and Attitudes during the MPox Outbreak: Findings from Topic Modeling of Tweets DOI Creative Commons
Nirmalya Thakur,

Yuvraj Nihal Duggal,

Zihui Liu

et al.

Computers, Journal Year: 2023, Volume and Issue: 12(10), P. 191 - 191

Published: Sept. 23, 2023

In the last decade and a half, world has experienced outbreaks of range viruses such as COVID-19, H1N1, flu, Ebola, Zika virus, Middle East Respiratory Syndrome (MERS), measles, West Nile just to name few. During these virus outbreaks, usage effectiveness social media platforms increased significantly, served virtual communities, enabling their users share exchange information, news, perspectives, opinions, ideas, comments related outbreaks. Analysis this Big Data conversations using concepts Natural Language Processing Topic Modeling attracted attention researchers from different disciplines Healthcare, Epidemiology, Science, Medicine, Computer Science. The recent outbreak MPox resulted in tremendous increase Twitter. Prior works area research have primarily focused on sentiment analysis content Tweets, few that topic modeling multiple limitations. This paper aims address gap makes two scientific contributions field. First, it presents results performing 601,432 Tweets about 2022 Mpox were posted Twitter between 7 May 3 March 2023. indicate during time may be broadly categorized into four distinct themes—Views Perspectives Mpox, Updates Cases Investigations LGBTQIA+ Community, COVID-19. Second, findings Tweets. show theme was most popular (in terms number posted) Views Mpox. followed by which themes COVID-19 respectively. Finally, comparison with studies is also presented highlight novelty significance work.

Language: Английский

Citations

5
Global recession sentiment analysis utilizing VADER and ensemble learning method with word embedding DOI Creative Commons
Maylinna Rahayu Ningsih, Kevyn Aalifian Hernanda Wibowo, Ahmad Ubai Dullah

et al.

Journal of Soft Computing Exploration, Journal Year: 2023, Volume and Issue: 4(3), P. 142 - 151

Published: Sept. 28, 2023

The issue of the Global Recession is hitting various countries, including Indonesia. Many Indonesians have expressed their opinions on global recession in 2023, one which from Twitter. By understanding public sentiment, we can assess impact felt by itself. Sentiment analysis this research a form support to evaluate Indonesia's sustainability dealing with accordance Sustainable Development Goals (SDGs). However, previous research, it still rare find model that has good performance conducting Analysis. Therefore, purpose propose machine learning expected provide sentiment analysis. existing dataset labeled Valence Aware Dictionary for Social Reasoning (VADER) algorithm, then an Ensemble Learning method designed composed Logistic Regression, Decision Tree, Random Forest, and Support Vector Machine (SVM) algorithms. After that, Countvectorizer feature extraction N-Gram, Best Match 25 (BM25), Word Embedding carried out convert sentences into numerical vectors so as improve performance. results more optimal accuracy 95.02% classifying sentiment. So proposed successfully performs better than research.

Language: Английский

Citations

5
Analyzing Public Reactions during the MPox Outbreak: Findings from Topic Modeling of Tweets DOI Open Access
Nirmalya Thakur,

Yuvraj Nihal Duggal,

Zihui Liu

et al.

Published: Sept. 1, 2023

In the last decade and a half, world has experienced outbreak of range viruses such as COVID-19, H1N1, flu, Ebola, Zika Virus, Middle East Respiratory Syndrome (MERS), Measles, West Nile just to name few. During these virus outbreaks, usage effectiveness social media platforms increased significantly served virtual communities, enabling their users share exchange information, news, perspectives, opinions, ideas, comments related outbreaks. Analysis this Big Data conversations outbreaks using concepts Natural Language Processing Topic Modeling attracted attention researchers from different disciplines Healthcare, Epidemiology, Science, Medicine, Computer Science. The recent MPox resulted in tremendous increase Twitter. Prior works field have primarily focused on sentiment analysis content Tweets, few that topic modeling multiple limitations. This paper aims address research gap makes two scientific contributions field. First, it presents results performing 601,432 Tweets about 2022 Mpox outbreak, which were posted Twitter between May 7, 2022, March 3, 2023. indicate during time may be broadly categorized into four distinct themes - Views Perspectives MPox, Updates Cases Investigations Mpox, LGBTQIA+ Community, COVID-19. Second, findings Tweets. show theme was most popular (in terms number posted) MPox. It is followed by COVID-19 respectively. Finally, comparison with prior also presented highlight novelty significance work.

Language: Английский

Citations

4