Social Media Sentiment Analysis Research Articles

Ensemble Classifiers for Arabic Sentiment Analysis of Social Network (Twitter Data) towards COVID-19-Related Conspiracy Theories

Sentiment analysis has recently become increasingly important with a massive increase in online content. It is associated with the analysis of textual data generated by social media that can be easily accessed, obtained, and analyzed. With the emergence of COVID-19, most published studies related to COVID-19’s conspiracy theories were surveys on the people's sentiments and opinions and studied the impact of the pandemic on their lives. Just a few studies utilized sentiment analysis of social media using a machine learning approach. These studies focused more on sentiment analysis of Twitter tweets in the English language and did not pay more attention to other languages such as Arabic. This study proposes a machine learning model to analyze the Arabic tweets from Twitter. In this model, we apply Word2Vec for word embedding which formed the main source of features. Two pretrained continuous bag-of-words (CBOW) models are investigated, and Naïve Bayes was used as a baseline classifier. Several single-based and ensemble-based machine learning classifiers have been used with and without SMOTE (synthetic minority oversampling technique). The experimental results show that applying word embedding with an ensemble and SMOTE achieved good improvement on average of F1 score compared to the baseline classifier and other classifiers (single-based and ensemble-based) without SMOTE.

Surprise, anticipation, sadness, and fear: A sentiment analysis of social media’s portrayal of police use of facial recognition technology

Abstract Policing agencies are adopting or trialling facial recognition technology (FRT). While the public tend to be sceptical of any new technology, public support is needed for both legitimacy and strong police–citizen relationships. The media can greatly influence not only the public agenda, but also the attitudes and sentiments towards the topic. This study takes an agenda-setting perspective to explore social media’s portrayal of police use of FRT. To do this, a sentiment analysis was conducted of 203 YouTube videos. Overall, the discourse was mostly positive for the use of FRT by police. An examination of the emotional language found high levels of surprise and anticipation along with sadness and fear. Notably, trust was expressed only in low levels. These findings inform the development of police practices and policies when adopting new technologies and the communication strategies of such policies and practices.

Sentiment Analysis in Social Media for Competitive Environment Using Content Analysis

Education sector has witnessed several changes in the recent past. These changes have forced private universities into fierce competition with each other to get more students enrolled. This competition has resulted in the adoption of marketing practices by private universities similar to commercial brands. To get competitive gain, universities must observe and examine the students’ feedback on their own social media sites along with the social media sites of their competitors. This study presents a novel framework which integrates numerous analytical approaches including statistical analysis, sentiment analysis, and text mining to accomplish a competitive analysis of social media sites of the universities. These techniques enable local universities to utilize social media for the identification of the most-discussed topics by students as well as based on the most unfavorable comments received, major areas for improvement. A comprehensive case study was conducted utilizing the proposed framework for competitive analysis of few top ranked international universities as well as local private universities in Lahore Pakistan. Experimental results show that diversity of shared content, frequency of posts, and schedule of updates, are the key areas for improvement for the local universities. Based on the competitive intelligence gained several recommendations are included in this paper that would enable local universities generally and Riphah international university (RIU) Lahore specifically to promote their brand and increase their attractiveness for potential students using social media and launch successful marketing campaigns targeting a large number of audiences at significantly reduced cost resulting in an increased number of enrolments.

Students’ Perceptions of E-Learning in Malaysian Universities: Sentiment Analysis Based Machine Learning Approach

Aim/Purpose: To gain insight into the opinions and reviews of Malaysian university students regarding e-learning systems, thereby improving the quality and services of these systems and resolving any problems, concerns, and issues that may exist within the institution. Background: This exploratory study examines the students’ perceptions of e-learning in Malaysia based on Sentiment Analysis (SA) to gain a clear insight into their feelings about the quality of e-learning systems and related services in Malaysian universities to determine whether these opinions are positive or negative. Methodology: The data was collected from Twitter; the Full Archive Search API Premium v1.1 tire was chosen to access the tweets from November 1, 2019, to December 30, 2020. The R programming language library package “rtweet” was applied to access the search API and query the tweets. To classify students’ opinions, sentiment analysis-based Machine Learning (ML) with Support Vector Machine (SVM) was utilized. Rapid Miner, a statistical and data mining tool, was used to determine the sentiment of tweets and the accuracy of the ML algorithm. After preparing the data, RapidMiner was used to pre-process and classify the final 1201 tweets based on sentiment, and National Research Council (NRC) word-emotion lexicon was used to detect the presence of eight emotions in the tweets. The confusion matrix is used to determine the classifier’s performance. Contribution: This research provided evidence for the effective use of sentiment analysis as an indicator that may contribute to the development of educational systems, specifically, e-learning systems in Malaysian universities. Findings: Based on the findings, the majority of students have a positive opinion about e-learning systems in Malaysian universities. Precisely, the results showed that 65% of sentiments were classified as positive and 35% as negative. Moreover, among the eight emotions, the majority of the tweets expressed a higher level of trust, anticipation, and joy. Recommendations for Practitioners: The study findings could help classify the teachers’ strengths and weaknesses graphically based on the students’ positive and negative feedback. These findings would also help decision-makers and educationalists be more aware of students’ feelings (sentiments) and concerns. Thus, using social media sentiment analysis should be encouraged as a valuable source of information that may assist their educational decision-making, e-learning development, and performance evaluation. Recommendation for Researchers: The findings may encourage other researchers to apply SA based ML approach and use Twitter as a data source to discover users’ opinions about certain issues in learning and teaching processes. Impact on Society: Our study confirmed that social media data could provide valuable and supportive information about educational systems and procedures in e-learning for appropriate decision-making regarding future development and strategies. Future Research: Future work can experiment with other classification models and different ML classification algorithms as well as other feature extraction methods and compare the results to find the best accuracy that can improve the classification results

Approximating Viewership of Streaming T.V Programs Using Social Media Sentiment Analysis

Approximating viewership of a program is essential for broadcast networks and cable (TNT, AMC, and HBO) as they profit by selling advertising space during programming (airing of the show). With the advent and increased popularity of streaming services such as Netflix and Amazon Prime Video, the model for providing television content to viewers has fundamentally changed. However, this change in the delivery model for television programs has also made it more challenging to determine the actual viewership of television programs on streaming services. Because television programs are no longer aired in pre-determined timeslots, viewership shares cannot be determined. To make matters even more complicated, streaming services like Netflix and Amazon Prime Video are proprietary platforms. They have traditionally kept viewership numbers closely guarded and have not released them to the public. This paper proposes a methodology to approximate the viewership of streaming television shows. This is achieved by exploiting data from social networks, i.e., publicly available information and sentiment score derived from sentiment analysis on tweets published about television programs, using Random Forest classifier. The proposed methodology achieved 85% accuracy in predicting the viewership of the streaming shows.

Exploring Consumer Emotions in Pre-Pandemic and Pandemic Times. A Sentiment Analysis of Perceptions in the Fine-Dining Restaurant Industry in Bucharest, Romania.

This research paper aims to analyse how consumer emotions have evolved during the pandemic period in comparison with the pre-pandemic period in relation to restaurant demand in the Romanian fine-dining industry and uses valuable information based on social-media sentiment analysis and content analysis. Focusing on theories of consumer behaviour, the study aims to emphasize how, under the influence of an epidemic crisis caused by an infectious disease, individual behaviour adapts to the “new normal”, embracing a series of changes in the preferences, attitudes, and cognitive choice-making processes. The article takes into account a comparative analysis of the consumer emotions between the pre-COVID-19 pandemic period (2010–2019) and the pandemic period (2020–present), based on the online reviews provided by customers for five fine-dining restaurants from Bucharest, the capital city of Romania: The Artist, Relais & Chateaux Le Bistrot Francais, Casa di David, Kaiamo, and L’Atelier. The research was based on two mining analyses—content analysis and sentiment analysis—and explored the emotional intent of words, with the data being collected from TripAdvisor through web-scrapping. The empirical results defined the fine-dining experience during the pandemic as being associated with the quality of the dishes and also with the quality of the service. The overall consumer sentiment in the direction of the restaurants analyzed is positive. The sentiment research found that throughout the epidemic, the consumers’ attitudes about restaurants deteriorated. In this sense, consumers seem to be less satisfied with the restaurants’ services than before the pandemic. This is another thing that the restaurants had difficulties in when adapting their operations for the pandemic.

Wearable IoT intelligent recommender framework for a smarter healthcare approach

PurposeThe increase in the number of healthcare wearable (Internet of Things) IoT options is making it difficult for individuals, healthcare experts and physicians to find the right smart device that best matches their requirements or treatments. The purpose of this research is to propose a framework for a recommender system to advise on the best device for the patient using machine learning algorithms and social media sentiment analysis. This approach will provide great value for patients, doctors, medical centers, and hospitals to enable them to provide the best advice and guidance in allocating the device for that particular time in the treatment process.Design/methodology/approachThis data-driven approach comprises multiple stages that lead to classifying the diseases that a patient is currently facing or is at risk of facing by using and comparing the results of various machine learning algorithms. Hereupon, the proposed recommender framework aggregates the specifications of wearable IoT devices along with the image of the wearable product, which is the extracted user perception shared on social media after applying sentiment analysis. Lastly, a proposed computation with the use of a genetic algorithm was used to compute all the collected data and to recommend the wearable IoT device recommendation for a patient.FindingsThe proposed conceptual framework illustrates how health record data, diseases, wearable devices, social media sentiment analysis and machine learning algorithms are interrelated to recommend the relevant wearable IoT devices for each patient. With the consultation of 15 physicians, each a specialist in their area, the proof-of-concept implementation result shows an accuracy rate of up to 95% using 17 settings of machine learning algorithms over multiple disease-detection stages. Social media sentiment analysis was computed at 76% accuracy. To reach the final optimized result for each patient, the proposed formula using a Genetic Algorithm has been tested and its results presented.Research limitations/implicationsThe research data were limited to recommendations for the best wearable devices for five types of patient diseases. The authors could not compare the results of this research with other studies because of the novelty of the proposed framework and, as such, the lack of available relevant research.Practical implicationsThe emerging trend of wearable IoT devices is having a significant impact on the lifestyle of people. The interest in healthcare and well-being is a major driver of this growth. This framework can help in accelerating the transformation of smart hospitals and can assist doctors in finding and suggesting the right wearable IoT for their patients smartly and efficiently during treatment for various diseases. Furthermore, wearable device manufacturers can also use the outcome of the proposed platform to develop personalized wearable devices for patients in the future.Originality/valueIn this study, by considering patient health, disease-detection algorithm, wearable and IoT social media sentiment analysis, and healthcare wearable device dataset, we were able to propose and test a framework for the intelligent recommendation of wearable and IoT devices helping healthcare professionals and patients find wearable devices with a better understanding of their demands and experiences.

Comparing Overall and Targeted Sentiments in Social Media during Crises

The tracking of citizens' reactions in social media during crises has attracted an increasing level of interest in the research community. In particular, sentiment analysis over social media posts can be regarded as a particularly useful tool, enabling civil protection and law enforcement agencies to more effectively respond during this type of situation. Prior work on sentiment analysis in social media during crises has applied well-known techniques for overall sentiment detection in posts. However, we argue that sentiment analysis of the overall post might not always be suitable, as it may miss the presence of more targeted sentiments, e.g. about the people and organizations involved (which we refer to as sentiment targets). Through a crowdsourcing study, we show that there are marked differences between the overall tweet sentiment and the sentiment expressed towards the subjects mentioned in tweets related to three crises events.

SentiXGboost: enhanced sentiment analysis in social media posts with ensemble XGBoost classifier

ABSTRACT Sentiment Analysis has emerged as one of the most challenging and significant research directions in text mining. The first milestone of Sentiment Analysis involves identifying the polarity of sentiments in social media and classifying them as positive or negative. Most of the previous work on sentiment analysis has focused on designing a classifier, for a specific dataset, using supervised machine learning approaches and feature extraction methods. Notably, choosing the most appropriate classification technique to correctly classify the sentiment polarity of text is crucial. Thus, this study proposes a novel ensemble classifier approach, which utilizes a combination of multiple feature sets with ensemble classification by combining multiple base classifiers, which are weak learners, into an ensemble classifier. These feature sets include Bag of Words, Term Frequency–Inverse Document Frequency, Part of Speech, N-gram, Opinion Lexicon, and Term Frequency. The experiments confirm that the proposed ensemble method outperforms all individual classifiers and significantly improves the overall sentiment classification performance on the most frequently used datasets in Sentiment Analysis.

A novel sentiment analysis framework for monitoring the evolving public opinion in real-time: Case study on climate change

Smart city analytics involves tracking, interpreting, and evaluating the sentiments and emotions that are shared via online social media channels. Sentiment analysis of social media posts has become increasingly prominent in recent years as a means of gaining insights into how members of the public perceive current affairs. The ongoing research in this domain has leveraged multiple types of sentiment analysis. However, although the existing approaches enable researchers to acquire retrospective insights into public opinion, they do not enable a real-time evaluation. In addition, they are not readily scalable and necessitate the analysis of a significant amount of posts (in the millions) to facilitate a more in-depth evaluation. The study outlined in this paper was designed to address these shortcomings by presenting a framework that facilitates a real-time evaluation of the evolution of opinions shared by prominent public features and their respective followers; that is, high-impact posts. The developed solution encompasses a sophisticated Bi-directional LSTM classifier that was implemented and tested using a dataset consisting of 278,000 tweets related to the topic of climate change. The outcomes reveal that the proposed classifier achieved the following accuracies: 88.41% for discrimination; 89.66% for anger; 87.01% for inspiration; and 87.52% for joy - with negative emotions being more accurately classified than positive emotions. Similarly, the sentiment classification performance yielded accuracies of 89.32% for support and 89.80% for strong support, as well as 88.14% for opposition and 87.52% for strong opposition. In addition, the findings of the study indicated that geographic location, chosen topic, cultural sensitivities, and posting frequency all play a critical role in public reactions to posts and the ensuing perspectives they adopt. The solution stands out from existing retrospective analysis methods because it does not rely on the analysis of vast quantities of data records; rather, it can perform real-time, high-impact content analysis in a resource-efficient and sustainable manner. This framework can be used to generate insights into how public opinion is developing on a real-time basis. As such, it can have meaningful application within social media analysis efforts.

Applying sentiment analytics to examine social media crises: a case study of United Airline's crisis in 2017

PurposeIn recent years, social media crises occurred more and more often, which negatively affect the reputations of individuals, businesses and communities. During each crisis, numerous users either participated in online discussion or widely spread crisis-related information to their friends and followers on social media. By applying sentiment analysis to study a social media crisis of airline carriers, the purpose of this research is to help companies take measure against social media crises.Design/methodology/approachThis study used sentiment analytics to examine a social media crisis related to airline carriers. The arousal, valence, negative, positive and eight emotional sentiments were applied to analyze social media data collected from Twitter.FindingsThis research study found that social media sentiment analysis is useful to monitor public reaction after a social media crisis arises. The sentiment results are able to reflect the development of social media crises quite well. Proper and timely response strategies to a crisis can mitigate the crisis through effective communication with the customers and the public.Originality/valueThis study used the Affective Norms of English Words (ANEW) dictionary to classify the words in social media data and assigned the words with two elements to measure the emotions: valence and arousal. The intensity of the sentiment determines the public reaction to a social media crisis. An opinion-oriented information system is proposed as a solution for resolving a social media crisis in the paper.

Addressing Biodisaster X Threats With Artificial Intelligence and 6G Technologies: Literature Review and Critical Insights.

BackgroundWith advances in science and technology, biotechnology is becoming more accessible to people of all demographics. These advances inevitably hold the promise to improve personal and population well-being and welfare substantially. It is paradoxical that while greater access to biotechnology on a population level has many advantages, it may also increase the likelihood and frequency of biodisasters due to accidental or malicious use. Similar to “Disease X” (describing unknown naturally emerging pathogenic diseases with a pandemic potential), we term this unknown risk from biotechnologies “Biodisaster X.” To date, no studies have examined the potential role of information technologies in preventing and mitigating Biodisaster X.ObjectiveThis study aimed to explore (1) what Biodisaster X might entail and (2) solutions that use artificial intelligence (AI) and emerging 6G technologies to help monitor and manage Biodisaster X threats.MethodsA review of the literature on applying AI and 6G technologies for monitoring and managing biodisasters was conducted on PubMed, using articles published from database inception through to November 16, 2020.ResultsOur findings show that Biodisaster X has the potential to upend lives and livelihoods and destroy economies, essentially posing a looming risk for civilizations worldwide. To shed light on Biodisaster X threats, we detailed effective AI and 6G-enabled strategies, ranging from natural language processing to deep learning–based image analysis to address issues ranging from early Biodisaster X detection (eg, identification of suspicious behaviors), remote design and development of pharmaceuticals (eg, treatment development), and public health interventions (eg, reactive shelter-at-home mandate enforcement), as well as disaster recovery (eg, sentiment analysis of social media posts to shed light on the public’s feelings and readiness for recovery building).ConclusionsBiodisaster X is a looming but avoidable catastrophe. Considering the potential human and economic consequences Biodisaster X could cause, actions that can effectively monitor and manage Biodisaster X threats must be taken promptly and proactively. Rather than solely depending on overstretched professional attention of health experts and government officials, it is perhaps more cost-effective and practical to deploy technology-based solutions to prevent and control Biodisaster X threats. This study discusses what Biodisaster X could entail and emphasizes the importance of monitoring and managing Biodisaster X threats by AI techniques and 6G technologies. Future studies could explore how the convergence of AI and 6G systems may further advance the preparedness for high-impact, less likely events beyond Biodisaster X.

Deep learning and multilingual sentiment analysis on social media data: An overview

Twenty-four studies on twenty-three distinct languages and eleven social media illustrate the steady interest in deep learning approaches for multilingual sentiment analysis of social media. We improve over previous reviews with wider coverage from 2017 to 2020 as well as a study focused on the underlying ideas and commonalities behind the different solutions to achieve multilingual sentiment analysis. Interesting findings of our research are (i) the shift of research interest to cross-lingual and code-switching approaches, (ii) the apparent stagnation of the less complex architectures derived from a backbone featuring an embedding layer, a feature extractor based on a single CNN or LSTM and a classifier, (iii) the lack of approaches tackling multilingual aspect-based sentiment analysis through deep learning, and, surprisingly, (iv) the lack of more complex architectures such as the transformers-based, despite results suggest the more difficult tasks requires more elaborated architectures.

Sentiment Polarity Classification at EVALITA: Lessons Learned and Open Challenges

Sentiment analysis in social media is a popular task attracting the interest of the research community, also in recent evaluation campaigns of natural language processing tasks in several languages. We report on our experience in the organization of SENTIment POLarity Classification Task (SENTIPOLC), a shared task on sentiment classification of Italian tweets, proposed for the first time in 2014 within the Evalita evaluation campaign. We present the datasets-which include an enriched annotation scheme for dealing with the impact of figurative language on polarity-the evaluation methodology, and discuss the approaches and results of participating systems. We also offer a reflection on the open challenges of state-of-the-art systems for sentiment analysis of microblogging in Italian, as they emerge from a qualitative analysis of misclassified tweets. Finally, we provide an evaluation of the resources we have created, and share the lessons learned by running this task for two consecutive editions.

Time Series for Forecasting Stock Market Prices Based on Sentiment Analysis of Social Media

This paper attempts to find a relation between the public perception of a company and its stock value price. Since social media is a very powerful tool used by a lot of people to voice their opinions on the performance of a company, it is a good source of information about the public sentiment. Previous studies have shown that the overall public sentiment collected from sites like Twitter do have a relation to the market price of a company over a period of time. The goal is to build on their research to improve the accuracy of predictions and determine if the public perception surrounding a company is a driving factor of its stock growth.

Optimize Machine Learning Programming Algorithms for Sentiment Analysis in Social Media

Optimize Machine Learning Programming Algorithms for Sentiment Analysis in Social Media

Social emotion classification of Japanese text information based on SVM and KNN

How to model multi-modal data and rich context information under the same model framework has become the key to user sentiment analysis in social media. This research uses SVM and KNN to build models for Japanese social sentiment classification and proposes a topic model for user sentiment analysis based on SVM and KNN. Moreover, taking TE process data as an example, this paper selects radial basis kernel function and grid search method for the construction of SVM classifier. In addition, by introducing the correlation variable between the sentiment of the comment and the sentiment of the tweet, the comment is related to the original tweet. Based on the establishment of the model, this research proposes a parameter estimation algorithm for model solving based on the idea of SVM-KNN, and uses the Twitter real data set as the experimental data set to verify the effectiveness of the user sentiment analysis model proposed in this paper. The research results show that the method proposed in this paper has a certain effect.

Tourism destination management using sentiment analysis and geo-location information: a deep learning approach

Sentiment analysis on social media such as Twitter is a challenging task given the data characteristics such as the length, spelling errors, abbreviations, and special characters. Social media sentiment analysis is also a fundamental issue with many applications. With particular regard of the tourism sector, where the characterization of fluxes is a vital issue, the sources of geotagged information have already proven to be promising for tourism-related geographic research. The paper introduces an approach to estimate the sentiment related to Cilento’s, a well known tourism venue in Southern Italy. A newly collected dataset of tweets related to tourism is at the base of our method. We aim at demonstrating and testing a deep learning social geodata framework to characterize spatial, temporal and demographic tourist flows across the vast of territory this rural touristic region and along its coasts. We have applied four specially trained Deep Neural Networks to identify and assess the sentiment, two word-level and two character-based, respectively. In contrast to many existing datasets, the actual sentiment carried by texts or hashtags is not automatically assessed in our approach. We manually annotated the whole set to get to a higher dataset quality in terms of accuracy, proving the effectiveness of our method. Moreover, the geographical coding labelling each information, allow for fitting the inferred sentiments with their geographical location, obtaining an even more nuanced content analysis of the semantic meaning.

The Performance of Indonesia’s President: A Sentiment Analysis in Social Media

The role of a leader is critical, especially when a person leads a country. Indonesia, as the biggest archipelago country, led by Jokowi since 2014. From his first period in 2014 until his second period started in 2019, many opinions, suggestions, and criticisms addressed to the president, Jokowi. This research aims to measure the performance of Jokowi as Indonesia’s president by collecting public opinions on social media, peculiarly in an interview video published on YouTube. All comments are scrapped and processed using machine learning tools, which are Python and TextBlob. The results of this research are positive and negative sentiments related to Jokowi. From the research conducted, the most influential positive sentiments of Jokowi are his personality. The negative sentiments gained in the video are about the poor quality of translation and the English skill of Jokowi.

Annotated Corpus of Mesopotamian-Iraqi Dialect for Sentiment Analysis in Social Media

Research on Sentiment Analysis in social media by using Mesopotamian-Iraqi Dialect (MID) of Arabic language was rarely found, there is no reliable dataset developed in MID neither an annotated corpus for the sentiment analysis of social media in this dialect. Therefore, this gap was the main stumbling block for researchers of sentiment analysis in MID, for this reason, this paper introduced the development of an annotated corpus of Mesopotamian-Iraqi Dialect for sentiment analysis in social media and named it as (ACMID) stands for (the annotated corpus of Mesopotamian-Iraqi Dialect) to help researchers in future for using this corpus for their studies, to the best of our knowledge this is the first annotated corpus that both classify polarity as well as emotion classification in MID. Likewise, Facebook as the most popular social platform among Iraqis was used to extract the data from its popular Iraqi pages. 5000 comments were extracted from these pages classified by its polarity (Positive, Negative, Neutral, Spam) by two Iraqi annotators, these annotators were simultaneously classifying the same comments according to Ekman seven universal emotions (Anger, Fear, Disgust, Happiness, Sadness, Surprise, Contempt) or no emotion. Cohen's kappa coefficient was then used to compare the two annotators’ results to find the reliability of these results. The data shows a comparable value among the two annotators for the polarity classification as high as 0.82, while for the emotion classification the result was 0.65.