VADER: A Parsimonious Rule-Based Model for Sentiment Analysis of Social Media Text

Sentiment analysis has applications in diverse contexts such as in the gathering and analysis of opinions of individuals about various products, issues, social, and political events. Understanding public opinion can help improve decision making. Opinion mining is a way of retrieving information via search engines, blogs, microblogs and social networks. Individual opinions are unique to each person, and Twitter tweets are an invaluable source of this type of data. However, the huge volume and unstructured nature of text/opinion data pose a challenge to analyzing the data efficiently. Accordingly, proficient algorithms/computational strategies are required for mining and condensing tweets as well as finding sentiment bearing words. Most existing computational methods/models/algorithms in the literature for identifying sentiments from such unstructured data rely on machine learning techniques with the bag-of-word approach as their basis. In this work, we use both unsupervised and supervised approaches on various datasets. Unsupervised approach is being used for the automatic identification of sentiment for tweets acquired from Twitter public domain. Different machine learning algorithms such as Multinomial Naive Bayes (MNB), Maximum Entropy and Support Vector Machines are applied for sentiment identification of tweets as well as to examine the effectiveness of various feature combinations. In our experiment on tweets, we achieve an accuracy of 80.68% using the proposed unsupervised approach, in comparison to the lexicon based approach (the latter gives an accuracy of 75.20%). In our experiments, the supervised approach where we combine unigram, bigram and Part-of-Speech as feature is efficient in finding emotion and sentiment of unstructured data. For short message services, using the unigram feature with MNB classifier allows us to achieve an accuracy of 67%.

With the advancement of sentiment analysis (SA) models and their incorporation into our daily lives, fairness testing on these models is crucial, since unfair decisions can cause discrimination to a large population. Nevertheless, some challenges in fairness testing include the unknown oracle, the difficulty in generating suitable test inputs, and the lack of a reliable way of fixing the issues. To fill in these gaps, BiasRV, a tool based on metamorphic testing (MT), was introduced and succeeded in uncovering fairness issues in a transformer-based model. However, the extent of unfairness in other SA models has not been thoroughly investigated. Our work conducts a more comprehensive empirical study to reveal the extent of fairness violations, specifically gender fairness, exhibited by other popular word embedding-based SA models. We define fairness violation as the behavior in which an SA model predicts variants created from a text, which merely differ in gender classes, to have different sentiments. Our inspection utilizing BiasRV uncovers at least 30 fairness violations (at BiasRV’s default threshold) in all three SA models. Realizing the importance of addressing such significant violations, we introduce adversarial patches (AP) as a way of patch generation in an automated program repair (APR) system to fix them. We adopt adversarial fine-tuning in AP by retraining SA models using adversarial examples, which are bias-uncovering test cases dynamically generated by a tool named BiasFinder at runtime. Evaluation of the SA models shows that our proposed AP reduces fairness violations by at least 25%.

Multilingual sentiment analysis poses significant challenges, especially in the context of languages with low resources. The study proposes a hybrid deep learning model based on the CNN-BiLSTM architecture to classify sentiment in multiple languages, including those with limited corpus and lexical resources. This model integrates the multilingual text representation of mBERT embedding with CNN's ability to extract local features and BiLSTM's power in capturing sequential contexts. Experiments were conducted on datasets that included various languages such as Indonesian, Hausa, Swahili, and Yoruba. The results of the evaluation showed that the proposed model achieved an accuracy of 84.3% and a macro F1 score of 83.1%, outperforming basic models such as Naive Bayes and independent BiLSTM. These findings suggest that the hybrid approach is effective in improving sentiment analysis performance across languages and has promising potential for real-world multilingual applications [1] T. I. Jain and D. Nemade, “Recognizing Contextual Polarity in Phrase-Level Sentiment Analysis,” Int. J. Comput. Appl., vol. 7, no. 5, pp. 12–21, 2010, doi: 10.5120/1160-1453. [2] A. Pak and P. Paroubek, “Twitter as a corpus for sentiment analysis and opinion mining,” Proc. 7th Int. Conf. Lang. Resour. Eval. Lr. 2010, pp. 1320–1326, 2010, doi: 10.17148/ijarcce.2016.51274. [3] I. Iin, R. Supriatna, M. Mulyawan, and D. Rohman, "The Application of Natural Language Processing in the Sentiment Analysis of the 2024 Vice Presidential Candidate Using the Naive Bayes Algorithm," JATI (Journal of Mhs. Tek. Inform., vol. 8, no. 1, pp. 1109–1115, 2024, doi: 10.36040/jati.v8i1.8572. [4] B. Ramadhani and R. R. Suryono, "Comparison of Naïve Bayes Algorithms and Logistic Regression for Metaverse Sentiment Analysis," J. Media Inform. Budidarma, vol. 8, no. 2, p. 714, 2024, doi: 10.30865/mib.v8i2.7458. [5] F. F. Mailoa, "Sentiment analysis of twitter data using text mining method on obesity problems in Indonesia," J. Inf. Syst. Public Heal., vol. 6, no. 1, p. 44, 2021, doi: 10.22146/jisph.44455. [6] E. Lutfina, W. Andriana, S. Quamila, P. Wiratmaja, and E. Febrianti, "Science, Technology and Management Journal Methods and Algorithms in Sentiment Analysis: Systematic Literature Review Info Articles," vol. 4, no. 2, pp. 67–79, 2024, [Online]. Available: http://journal.unkartur.ac.id/index.php/stmj [7] A. Fauzi, M. F. Akbar, and Y. F. A. Asmawan, "Sentiment of Internet Analysis on Social Media Using Bayes Algorithm," J. Inform., vol. 6, no. 1, pp. 77–83, 2019, doi: 10.31311/ji.v6i1.5437. [8] D. Winoto, V. Desta Aditia, C. Sorisa, R. Priskila, and V. Handrianus Pranatawijaya, "Sentiment Analysis on User Reviews of Duolingo Language Learning Application: Using Naïve Bayes and K-Nearest Neighbor Algorithms," JATI (Journal of Mhs. Tek. Inform., vol. 8, no. 3, pp. 3230–3236, 2024, doi: 10.36040/jati.v8i3.9647. [9] T. Y. Pahtoni and H. Jati, "Analysis of Twitter Data Sentiment Related to ChatGPT Using Orange Data Mining," J. Techno. Inf. and Computing Science., vol. 11, no. 2, pp. 329–336, 2024, doi: 10.25126/jtiik.20241127276. [10] A. Ardiansyah, E. Argarini Pratama, N. Imam Fadlilah, and U. Bina Sarana Informatika, "Analysis of User Sentiment Towards the ChatGPT Application on the Google Play Store: The Application of the Support Vector Machine Algorithm," vol. 11, no. 2, pp. 247–254, 2024. [11] Normah, B. Rifai, S. Vambudi, and R. Maulana, "Sentiment Analysis of Vtuber Development Using SMOTE-Based Support Vector Machine Method," J. Tek. Computer. AMIK BSI, vol. 8, no. 2, pp. 174–180, 2022, doi: 10.31294/jtk.v4i2. [12] S. F. Intan, I. Permana, F. N. Salisah, M. Afdal, and F. Muttakin, "Comparison of KNN, NBC, and SVM Algorithms: Analysis of Public Sentiment Towards Parking in the City of Pekanbaru," JUSIFO (Journal of Sist. Information), vol. 9, no. 2, pp. 85–96, 2023, doi: 10.19109/jusifo.v9i2.21357. [13] M. A. Maulana, A. Setyanto, and M. P. Kurniawan, "Analysis of Social Media Sentiment at Amikom University Yogyakarta as a Means of Information Dissemination Using the Svm Classification Algorithm," Sem. Nas. Technology. Inf. and Multimed. 2018 Univ. AMIKOM Yogyakarta, 10 February 2018Pp. 7–12, 2018. [14] D. A. Efraim, "Sentiment Analysis on Instagram Social Media Using Naive Bayes Algorithm (Case Study: Indonesian Futsal National Team)," no. April 2012, pp. 498–509, 2023. [15] I. Maulana, W. Apriandari, and A. Pambudi, "Aspect-Based Sentiment Analysis of Mypertamina Application Reviews Using Support Vector Machine," IDEALIS Indones. J. Inf. Syst., vol. 6, no. 2, pp. 172–181, 2023, doi: 10.36080/idealis.v6i2.3022. [16] N. D. Putranti and E. Winarko, "Twitter Sentiment Analysis for Indonesian Text with Maximum Entropy and Support Vector Machine," IJCCS (Indonesian J. Comput. Cybern. Syst., vol. 8, no. 1, p. 91, 2014, doi: 10.22146/ijccs.3499. [17] R. Maulana, A. Voutama, and T. Ridwan, "Sentiment Analysis of MyPertamina Application Reviews on Google Play Store using NBC Algorithm," J. Techno. Integrated, vol. 9, no. 1, pp. 42–48, 2023, doi: 10.54914/jtt.v9i1.609. [18] M. Y. Pratama, U. A. Putri, P. A. D. Angraini, D. Puspita, and F. Kurniawan, "Sentiment Analysis of Chat GPT as the Future of Workers on Youtube Social Media using the Naive Bayes Classification Algorithm," Explore. J. Sist. Inf. and Telemat., vol. 14, no. 2, p. 193, 2023, doi: 10.36448/jsit.v14i2.3391.

Deep Learning Models Based Sentiment Analysis Application of multiple layers of artificial neural networks for the learning tasks is called deep learning. In the research field, deep learning is a powerful machine learning technique. It has the ability to learn multiple levels of representations and abstractions from data, which can solve both supervised and unsupervised learning tasks. Deep learning uses multiple layers of non-linear processing units for feature extraction and classification. Sentiment analysis is one of the active research areas in Natural Language Processing. There exist numerous techniques to perform sentiment analysis task, which include both supervised and unsupervised methods. Types of supervised machine learning method include Support Vector Machines (SVM), Maximum Entropy, Nave Bayes, etc. Types of unsupervised machine learning methods include sentiment lexicons, grammatical analysis, and syntactic patterns. Application of deep learning to sentiment analysis has been very popular now a days. The reason to choose deep learning models, as it provides improved performance and accurate results over learning tasks. Deep neural network methods will perform both feature extraction and classification for document and short text. The application of different deep learning models on sentiment analysis. Sentiment Analysis Sentiment analysis is a technique that comes under the field of natural language processing. The process of identifying human emotions and thinking is termed as sentiment analysis, which is also known as opinion mining. It classifies whether the given text is positive or negative or sometimes neutral also, based on the classification level on a given document or sentence. There exist several approaches to accomplish the sentiment analysis task. This task is achieved by identifying the sentiment or opinion of the subjective element within a text. The approaches that are used to classify a piece of text are according to the opinions expressed in it, i.e. either positive or negative or neutral. The analyzing piece of text can be sentence or document or anything. The sentiment analysis is accomplished by classifying the methods into machine learning and lexicon-based approach. Again the machine learning approach is classified into supervised and unsupervised machine learning techniques. Under supervised learning there exists mainly Support Vector Machine (SVM), Neural Networks (NN), Nave Bayes (NB), Maximum Entropy (ME) approaches.

The key issue at this stage is how to mine the large amount of valuable user sentiment information from the massive amount of web text and create a suitable dynamic user text sentiment analysis technique. Hence, this study offers a writing feature abstraction process based on ON-LSTM and attention mechanism to address the problem that syntactic information is ignored in emotional text feature extraction. The study found that the Att-ON-LSTM improved the micro-average F1 value by 2.27% and the macro-average F value by 1.7% compared to the Bi-LSTM model with the added attentivity mechanisms. It is demonstrated that it can perform better extraction of semantic information and hierarchical structure information in emotional text and obtain more comprehensive emotional text features. In addition, the ON-LSTM-LS, a sentiment analysis model based on ON-LSTM and tag semantics, is planned to address the problem that tag semantics is ignored in the process of text sentiment analysis. The experimental consequences exposed that the accuracy of the ON-LSTM and labeled semantic sentiment analysis model on the test set is improved by 0.78% with the addition of labeled word directions compared to the model Att-ON-LSTM without the addition of labeled semantic information. The macro-averaged F1 value improved by 1.04%, which indicates that the sentiment analysis process based on ON-LSTM and tag semantics can effectively perform the text sentiment analysis task and improve the sentiment classification effect to some extent. In conclusion, deep learning models for dynamic user sentiment analysis possess high application capabilities.

The complexity of the Arabic language in terms of morphology, orthography, and dialects renders it more difficult to conduct sentiment analysis for the Arabic language. This issue is made much more challenging by the practice of extracting text features from short communications to evaluate the tone of the communication. On the other hand, the technique for analyzing and assessing sentiment faces a great deal of difficulty. These issues might be hampered by the accurate interpretation of sentiments and identifying the appropriate polarity of sentiment. Sentiment analysis can recognize and extract subjective information from the text. This study intends to investigate the effectiveness of various Machine Learning (ML) techniques in understanding the sentiments conveyed by the Arabic language. In addition, the feature extraction from the dataset was carried out with the help of the Term Frequency-Inverse Document Frequency (TF-IDF). As a consequence of this, the techniques of Adaboost classifier (AC), Support Vector Machine (SVM), Maximum Entropy (ME), Decision tree (DT), and K-Nearest Neighbors (KNN) are utilized in the process of sentiment analysis (SA). In conclusion, a model for ensemble-based sentiment analysis was developed. Compared to other machine learning classifiers cited earlier, we achieved better performance in terms of accuracy, precision, kappa, and ROC AUC-score for ensemble classifiers with 10-fold cross-validation.

Sentiment analysis of online user generated content is important for many social media analytics tasks. Researchers have largely relied on textual sentiment analysis to develop systems to predict political elections, measure economic indicators, and so on. Recently, social media users are increasingly using additional images and videos to express their opinions and share their experiences. Sentiment analysis of such large-scale textual and visual content can help better extract user sentiments toward events or topics. Motivated by the needs to leverage large-scale social multimedia content for sentiment analysis, we utilize both the state-of-the-art visual and textual sentiment analysis techniques for joint visual-textual sentiment analysis. We first fine-tune a convolutional neural network (CNN) for image sentiment analysis and train a paragraph vector model for textual sentiment analysis. We have conducted extensive experiments on both machine weakly labeled and manually labeled image tweets. The results show that joint visual-textual features can achieve the state-of-the-art performance than textual and visual sentiment analysis algorithms alone.

Sentiment analysis of online user generated content is important for many social media analytics tasks. Researchers have largely relied on textual sentiment analysis to develop systems to predict political elections, measure economic indicators, and so on. Recently, social media users are increasingly using additional images and videos to express their opinions and share their experiences. Sentiment analysis of such large-scale textual and visual content can help better extract user sentiments toward events or topics. Motivated by the needs to leverage large-scale social multimedia content for sentiment analysis, we propose a cross-modality consistent regression (CCR) model, which is able to utilize both the state-of-the-art visual and textual sentiment analysis techniques. We first fine-tune a convolutional neural network (CNN) for image sentiment analysis and train a paragraph vector model for textual sentiment analysis. On top of them, we train our multi-modality regression model. We use sentimental queries to obtain half a million training samples from Getty Images. We have conducted extensive experiments on both machine weakly labeled and manually labeled image tweets. The results show that the proposed model can achieve better performance than the state-of-the-art textual and visual sentiment analysis algorithms alone.

Sentiment analysis for Indonesian social media text is very important because the text content in social media is very diverse and requires an accurate method that can produce an analysis describing the state of the actual data. The main problem in sentiment analysis for Indonesian text on social media is unstructured text data and the use of non-standard languages such that sentiment analysis often produces errors. This paper focuses on sentiment analysis using a hybrid model that combines lexicon based and maximum entropy methods to classify the sentiments of Indonesian public opinion on government. The method consists of extracting datasets, preprocessing, lexicon-based classification, machine learning training, machine learning classification, and result interpretation. The results of the study produce 91 classifications of neutral sentiment, 51 document negative sentiments, 39 document positive sentiments and 152 document of mix sentiments. Based on the evaluation results, the hybrid sentiment model for Indonesian Language sentiment analysis on social media produced a pretty good accuracy score of 84.31% compared to previous studies. The implication of this study is to produce a sentiment analysis system with a hybrid method for Indonesian text on social media.

To study the application of convolutional neural networks (CNN) in microblog sentiment analysis, a microblog sentiment dictionary is established first. Then, latent Dirichlet allocation (LDA) is proposed for user forwarding sentiment analysis. The sentiment analysis models of CNN and long short-term memory network (LSTM) are established. Experiments are conducted to verify the application effect. The main contributions of this work encompass the establishment of a sentiment lexicon for Weibo, the optimization of two sentiment analysis models, namely CNN and LSTM, as well as the comparison and analysis of the performance of three sentiment analysis approaches: CNN, LSTM, and LDA. The research findings indicate that the CNN model achieves a prediction accuracy of 78.6% and an actual output precision of 79.3%, while the LSTM model attains a prediction accuracy of 83.9% and an actual output precision of 84.9%. The three analysis models all have high sentiment analysis accuracy. Among them, LDA analysis model has the advantages of universality and irreplaceable in text classification, while LSTM analysis model has relatively higher accuracy in sentiment analysis of users forwarding microblog. In short, each sentiment analysis model has its own strengths, and reasonable allocation and use can better classify microblog sentiment.

Social media has been a real-world sensor to observe the pulse of society. Although it provides unique communication opportunities, it also brings along vital challenges. One of them is hate speech, which attacks a single individual or targeted groups. Previous researchers claim that among social networks, the Twitter platform is mostly used to spread hate speech. However, data on a larger scale makes it hard to capture and understand the nature of hate speech on Twitter within specific food brands. In this study, sentiment analysis techniques were used to filter hate speech on Twitter and on three popular food brands in Malaysia. This study was conducted in five phases, namely raw data collection, pre-processing, sentiment analysis, visualization, and performance evaluation. This corpus consists of 28,898 data samples based on user tweet searches. A Twitter API was created and SQLite was used to store all the sample data. VADER sentiment analysis is used to classify tweet sentiment into positive, negative, and neutral. In the visualization phase, these three food brands are visualized using a histogram to gain sentiment analysis insights. Then, three machine learning methods were implemented to predict the best model for sentiment analysis. The Decision Tree classifier outperforms the average accuracy in Support Vector Machine and Logistic Regression with 99.99% for the training data set. This study provides insights to assist humans in making decisions. With the growth of opinions expressed in multimedia on social media, such as spoken feedback on Twitter, sentiment analysis has the potential to become a more news aggregation and low-cost endeavour.

In this paper, Maximum Entropy (ME) framework is used to classify text documents. The ME framework has a lot of advantages when compared with other supervised learning algorithms, such as naive Bayes classifier. For example, it makes no inherent conditional independence assumptions between terms. With four labeled data sets, extensive experiments are made to compare the accuracy of ME algorithm with those of naive Bayes and Support Vector Machine (SVM), which are two popular and accurate algorithms in the domain of text classification. The final result is that ME method consistently outperforms naive Bayes and SVM algorithms in accuracy. On the WebKB and Industry Vector data sets, the accuracy of ME algorithm increases from 81.38% to 85.52% and from 85.73% to 89.78% respectively. On the third 20 Newsgroups data set, our experimental result is opposite to that of Nigam et al. For the last Reuters-21578 data set, the accuracy of ME algorithm increases from 94.76% to 96.16%.

Sentiment Analysis is a branch of Natural Language Processing which intends to identify the sentiment in the data being analyzed through polarity analysis and emotion analysis. It can be performed with approaches which are based on Machine Learning as well as the approaches which are based on Lexicons which in turn rely on corpus and dictionaries. Thus, the unprecedented growth of E-Commerce Platforms resulting in an exponential increase in the amount of customer reviews prompts us to choose the most appropriate and optimized models of Sentiment Analysis to produce high accuracy. Since the customer reviews are one of the most important factors which influence the brand value, advertising and customer services of a company, harnessing Sentiment Analysis to get more insight into the reviews is the need of the hour. In this paper, the Customer Review Sentiment Analysis for polarity classification has been performed using Support Vector Machine Model and Convolutional Neural Network Model on a real world dataset of web scraped customer reviews, following which the Support Vector Machine Model was deployed to a web application. By choosing the most appropriate methods of dataset cleaning, text preprocessing and hyper parameter tuning, the Support Vector Machine and the Convolutional Neural Network Model achieved high accuracies of 96% and 94% respectively. Thus, both the Sentiment Analysis models have achieved much higher accuracy and minimal error rate in contrast to the existing models.

Virtual reality (VR) technology within the hotel industry marks a transformative shift in the way guests experience and engage with hospitality services. Virtual reality, with its immersive and interactive capabilities, enables hotels to provide a novel and engaging environment for guests. From virtual tours of hotel rooms and amenities to immersive experiences showcasing local attractions and cultural highlights, VR has the potential to revolutionize the pre-booking and on-site guest experience. This paper focused on the user experiences within hotel rooms enhanced with virtual reality (VR) technology. Leveraging content analysis, sentiment analysis, and advanced classification models, we aim to unravel the intricacies of user sentiments and preferences in this evolving domain. The content analysis reveals a spectrum of user opinions, ranging from enthusiastic endorsements of immersive VR content to nuanced critiques of room ambiance and interactivity. Subsequently, a sentiment analysis model accurately categorizes these sentiments, showcasing its effectiveness in capturing the diverse user expressions. Our classification analysis demonstrates the robustness of the sentiment analysis model, with high accuracy, precision, recall, and F1-score metrics. Comparatively, we introduce a proposed BERT model, harnessing advanced natural language processing techniques, and observe its performance against traditional sentiment analysis and an AutoEncoder Model. The results indicate that the BERT model matches the performance of traditional sentiment analysis, outperforming the AutoEncoder Model. This underscores the effectiveness of leveraging state-of-the-art language models in understanding and classifying user sentiments.

Virtual reality (VR) technology within the hotel industry marks a transformative shift in the way guests experience and engage with hospitality services. Virtual reality, with its immersive and interactive capabilities, enables hotels to provide a novel and engaging environment for guests. From virtual tours of hotel rooms and amenities to immersive experiences showcasing local attractions and cultural highlights, VR has the potential to revolutionize the pre-booking and on-site guest experience. This paper focused on the user experiences within hotel rooms enhanced with virtual reality (VR) technology. Leveraging content analysis, sentiment analysis, and advanced classification models, we aim to unravel the intricacies of user sentiments and preferences in this evolving domain. The content analysis reveals a spectrum of user opinions, ranging from enthusiastic endorsements of immersive VR content to nuanced critiques of room ambiance and interactivity. Subsequently, a sentiment analysis model accurately categorizes these sentiments, showcasing its effectiveness in capturing the diverse user expressions. Our classification analysis demonstrates the robustness of the sentiment analysis model, with high accuracy, precision, recall, and F1-score metrics. Comparatively, we introduce a proposed BERT model, harnessing advanced natural language processing techniques, and observe its performance against traditional sentiment analysis and an AutoEncoder Model. The results indicate that the BERT model matches the performance of traditional sentiment analysis, outperforming the AutoEncoder Model. This underscores the effectiveness of leveraging state-of-the-art language models in understanding and classifying user sentiments.