Sentiment Analysis On Movie Streaming Services On Twitter Using LDA and SVM Methods

Technological developments provide significant changes in the field of journalism, especially in the news by the mass media. The online news portal currently provides a comment column feature for news readers to comment on the existing news. This study aims to (1) look at the effect of reader comments on students' perceptions of controversial news about campus through the LINE Today feature, (2) analyze the difference in influence between positive comments and negative comments on students' perceptions of campus controversy news through the LINE Today feature. This study uses a quantitative approach with a questionnaire method to the research population. The sample is students of the Department of Communication, Faculty of Social and Political Sciences, Hasanuddin University who have accounts on the LINE application, totaling 199 people with the total sampling method. Regression analysis technique was used to see the effect of reader comments on student perceptions. To compare the effect of positive and negative reader comments, a comparative analysis technique is used, namely variance. The results showed that (1) There was a significant effect of reader comments on students' perceptions of campus controversy news through the LINE Today feature using both positive and negative stimuli; (2) There is a difference in the effect of positive and negative comments on students' perceptions of campus controversy news through the LINE Today feature.stimuli have a greater influence on student perceptions

Social media has become a one-stop shop for consuming news and expressing political views. Politics has become increasingly emotional, and the ensuing polarization has created echo chambers that favor narratives and stories that repeat only one point of view. In this article, we investigated the role of political activity through Twitter (now ‘X’) engagement as a predictor of destructive fires and deforestation in the Brazilian Legal Amazon (BLA). We used a machine learning approach based on sentiment analysis and satellite data. To test the consistency of the sentiment analysis, we compared the timing of messages related to fire and deforestation events with daily fire data from satellites. When comparing positive and negative comments about fires in the BLA, the results showed that the best model for predicting fire outbreaks is the decision tree regressor. We found evidence that positive comments about agriculture, industry, and the Amazon rainforest in response to speeches and statements by high-ranking Brazilian politicians tend to induce positive comments about fire outbreaks and deforestation. These comments then become good predictors of fire outbreaks with a 6-day lag. These results support the view that high-ranking politicians have enormous power to influence damaging events that can have severe impacts on communities, the environment, and the economy. Brazil has seen an unprecedented increase in deforestation and fires in the Amazon rainforest in recent years. Our findings contribute to the growing literature on the role of social media in real-world events and how machine learning approaches can be used to address this class of problems.

Category:Practice ManagementIntroduction/Purpose:Utilization by patients of physician rating websites continues to expand. There is limited information on how these websites function and influence patient perception and physicians’ practices. No study has specifically investigated online ratings and comments of orthopedic foot and ankle surgeons. In this study, we identified what factors impact online ratings and comments for this subset of surgeons.Methods:A total of 210 orthopedic foot and ankle surgeons in or near metropolitan areas were randomly selected from the American Orthopaedic of Foot & Ankle Society (AOFAS) website. Demographic information, ratings (number of stars), and comments were reviewed on the three most visited public domain physician ratings websites: HealthGrades.com (HealthGrades), Vitals.com (Vitals), and Ratemds.com (Ratemds). Content differences between positive and negative comments were evaluated.Results:Orthopedic foot and ankle surgeons had a mean rating of 4.03 ± 0.57 out of 5 stars. A high percentage (84%) of the total number of ratings were either 1-star (17%) or 5-stars (67%). Most positive comments related to outcome, physician personality, and communication while most negative comments related to outcome, bedside manner, and waiting time. Chi-square analysis revealed statistically significant proportions of positive comments pertaining to surgeon-dependent factors (eg, physician personality, knowledge, skills) and of negative comments concerning surgeon-independent factors (eg, waiting time, logistics).Conclusion:This is the first study examining online ratings and written comments of orthopedic foot and ankle surgeons. Surgeons had a generally favorable rating and were likely to have positive comments. Patients were likely to write positive comments about surgeon personality and communication, and negative comments pertaining to bedside manner and waiting time. Knowledge and management of online content may allow surgeons to improve patient satisfaction and online ratings.

In the year 2018, 18.9% of the population in Indonesia mentioned that the main reason for their use of the Internet is social media. One of the social media with an active user of 6.43 million users is Twitter. Based on the surge of information published via Twitter, it is possible that such information may contain the user's opinions on an object, such objects may be events around the community such as a product or service. This makes the company use Twitter as a medium to disseminate information. An example is an Internet Service Provider (ISP) such as Indihome. Through Twitter, users can discuss each other's complaints or satisfaction with Indihome's services. It takes a method of sentiment analysis to understand whether the textual data includes negative opinions or positive opinions. Thus, the authors use the Support Vector Machine (SVM) method in sentiment analysis on the opinions of the Indihome service user on Twitter, with the aim of obtaining a sentiment classification model using SVM, and to know how much accuracy the SVM method generates, which is applied to sentiment analysis, and to see how satisfied the Indihome service users are based on Twitter. After testing with SVM method The result is accuracy 87%, precision 86%, recall 95%, error rate 13%, and F1-score 90%

ObjectivesPrior research has demonstrated that men and women emergency medicine (EM) residents receive similar numerical evaluations at the beginning of residency, but that women receive significantly lower scores than men in their final year. To better understand the emergence of this gender gap in evaluations we examined discrepancies between numerical scores and the sentiment of attached textual comments.MethodsThis multicenter, longitudinal, retrospective cohort study took place at four geographically diverse academic EM training programs across the United States from July 1, 2013–July 1, 2015 using a real-time, mobile-based, direct-observation evaluation tool. We used complementary quantitative and qualitative methods to analyze 11,845 combined numerical and textual evaluations made by 151 attending physicians (94 men and 57 women) during real-time, direct observations of 202 residents (135 men and 67 women).ResultsNumerical scores were more strongly positively correlated with positive sentiment of the textual comment for men (r = 0.38, P < 0.001) compared to women (r = −0.26, P < 0.04); more strongly negatively correlated with mixed (r = −0.39, P < 0.001) and negative (r = −0.46, P < 0.001) sentiment for men compared to women (r = −0.13, P < 0.28) for mixed sentiment (r = −0.22, P < 0.08) for negative; and women were around 11% more likely to receive positive comments alongside lower scores, and negative or mixed comments alongside higher scores. Additionally, on average, men received slightly more positive comments in postgraduate year (PGY)-3 than in PGY-1 and fewer mixed and negative comments, while women received fewer positive and negative comments in PGY-3 than PGY-1 and almost the same number of mixed comments.ConclusionWomen EM residents received more inconsistent evaluations than men EM residents at two levels: 1) inconsistency between numerical scores and sentiment of textual comments; and 2) inconsistency in the expected career trajectory of improvement over time. These findings reveal gender inequality in how attendings evaluate residents and suggest that attendings should be trained to provide all residents with feedback that is clear, consistent, and helpful, regardless of resident gender.

‘I wish my body looked like theirs!’: How positive appearance comments on social media impact adolescents' body dissatisfaction

One of the social media with 24.85 million active users is Twitter. Information published on Twitter can be in the form of user opinions on an object, such as a product or service. Therefore, the company utilizes Twitter as a medium to disseminate information. This makes the company use Twitter as a medium to disseminate information. An example is an Internet Service Provider (ISP) company such as Indihome. Through Twitter, users can discuss their complaints and satisfaction with Indihome services. A method is needed, namely sentiment analysis to understand whether the textual data includes neutral opinion, negative opinion or positive opinion. So, the authors use the Support Vector Machine (SVM) method in sentiment analysis of Indihome service user opinions on Twitter, with the aim of getting a sentiment classification model using SVM and to find out how much accuracy is produced by the SVM method applied to sentiment analysis and to find out how satisfied Indihome service users are based on Twitter. After testing with the SVM method the results are 91% accuracy. Precision 51% Recall 75% and F1-Score 59%.

Technological developments are increasingly rapid and moving towards digital, which in the end technology can also help people who are experiencing economic problems, namely with online loan services. Even though there are many conveniences provided by online loan services, of course not all people give positive comments because there are quite a few negative comments about this service.One of the social media that is widely used by the public to provide comments about online loans is Twitter. Sentiment analysis is a data processing process to obtain information about whether an opinion sentence tends to be positive, negative or even neutral. This research contains sentiment analysis towards Online Loans on Twitter using the Lexicon Based and Support Vector Machine methods. From the results of this research, the accuracy for SVM was 82.36%. From these results it can be concluded that the use of the Lexicon Based and Support Vector Machine methods is considered quite good and effective for classifying sentiment

All groups of people felt the impact of the COVID-19 pandemic. This situation triggers anxiety, which is bad for everyone. The government's role is very influential in solving these problems with its work program. It also has many pros and cons that cause public anxiety. For that, it is necessary to detect anxiety to improve government programs that can increase public expectations. This study applies machine learning to detecting anxiety based on social media comments regarding government programs to deal with this pandemic. This concept will adopt a sentiment analysis in detecting anxiety based on positive and negative comments from netizens. The machine learning methods implemented include K-NN, Bernoulli, Decision Tree Classifier, Support Vector Classifier, Random Forest, and XG-boost. The data sample used is the result of crawling YouTube comments. The data used amounted to 4862 comments consisting of negative and positive data with 3211 and 1651. Negative data identify anxiety, while positive data identifies hope (not anxious). Machine learning is processed based on feature extraction of count-vectorization and TF-IDF. The results showed that the sentiment data amounted to 3889 and 973 in testing, and training with the greatest accuracy was the random forest with feature extraction of vectorization count and TF-IDF of 84.99% and 82.63%, respectively. The best precision test is K-NN, while the best recall is XG-Boost. Thus, Random Forest is the best accurate to detect someone's anxiety based-on data from social media.

PurposeThe emergence of social media has provided an alternative mechanism for students to provide and access online reviews about universities. These forms of feedback are often not systematically monitored and interpreted by universities. The purpose of this paper is to analyse quantitative ratings and qualitative comments provided by students via Google reviews of 40 Australian universities. The qualitative comments were analysed via thematic coding which identified four key themes including quality of teaching, course design, learning environment and administrative support. The data were also subjected to automated content analysis using the text analytics software Leximancer to enhance the objectivity and reliability of findings.Design/methodology/approachThe study involved the extraction of quantitative ratings as well as qualitative comments from Google reviews of 40 Australian universities. Each university was searched on Google and their rating and comments were collected as of August 2019. The qualitative data were included in an Excel spreadsheet along with the name of the universities, and their quantitative ratings. Using thematic coding techniques, the qualitative data were manually coded into four themes: quality of teaching, course design, learning environment and administrative support. Furthermore, the qualitative comments were grouped into two categories: positive comments and negative comments.FindingsThe paper argues the need for universities to actively monitor these new forms of student feedback which have provided autonomy and freedom for students and other stakeholders to access and provide 24/7 feedback. The study found common themes emerging from the positive and negative student comments related to: quality of teaching, course design, learning environments and administrative support. These aspects contributed to the students’ satisfaction levels when they reflected on their educational experience on Google reviews. The study also found that institutions with high numbers of negative comments tend to have lower ratings compared to those with positive comments.Research limitations/implicationsThe study did not use other learning analytic tools to measure and assess the feedback from students using social media and other forms of feedback.Originality/valueMany studies have been undertaken on student experience. Studies have focussed on the experience of home and international students, on-campus and online students, and students from various underrepresented backgrounds. So far no study has been undertaken on the reviews provided by students using Google reviews.

The aim of the research is to identify the representations of the Yemeni reality in virtual societies in the conditions of war, through interactions with viewers, icons, and positive and negative comments with static images, which were displayed through the “Create a Story” window on the researcher’s page on the social networking platform (Facebook) to be the research community and its sample. The analytical descriptive approach was used in analyzing the interactions, and the number of images reached (60) static images during the extended period (1/1 - 21/3/ 2021), each four images were distributed in each of the topics of the Yemeni reality, and the most prominent findings of the research : The images of the topics of prominent Yemeni feminist elites, Yemeni women in traditional costumes and actors are given priority by representations of the Yemeni reality, through iconic interactions, with a high rate of (32.21%), and through interaction by viewing, at a high rate of (30.20%), as these topics came In the forefront of positive representations through interaction with positive comments, and with them pictures of common folk dance topics (men, women), scholars and contemporary thinkers, poets and writers, with a percentage of (73.21%) of the high representations. It did not show any positive high representations in the images of political, economic, scientific, intellectual, literary, artistic and civilized subjects. While the results showed the predominance of medium representations, through the interaction by watching the images of the warring Yemeni political parties, the Yemeni Congregation for Reform Party topped the negative representations, through the iconic interactions, then the topics of merchants, businessmen, the legitimate authority and the followers of former President Ali Abdullah Saleh, and issues of public opinion and the destruction that He was succeeded by the war and the Houthis (Ansar Allah), with a percentage of (78.70%) of the average representation.

The rapid growth of social media provides businesses with a new venue for users to make a purchase, as well as the opportunity to post their reviews and comments on products, which can greatly influence the purchase decision process of potential users. The social media platform generates both positive and negative reviews and comments that can influence users' purchase behaviour. In order to gain a deeper understanding of consumer purchase behaviour on social media, the study examines how positive and negative reviews and comments can affect purchase attitudes and pre-purchase behaviour. It also examines how social media users' purchase attitudes contribute to their specific purchase behaviours such as pre-purchase and post-purchase behaviours. The research instrument was a questionnaire, and data were collected through social media platforms using Google forms. Descriptive statistics and Spearman correlation coefficients were used to achieve the objective of this study. The research sample consisted of 162 social media users. The results suggest that there is a positive relationship between purchase attitude and pre-purchase and post-purchase behaviour. The results also suggest that there is a positive relationship between positive and negative reviews and comments, purchase attitude, and pre-purchase behaviour.

Twitter has provided a convenient platform to express feelings and opinions in different areas. Opinion mining in Twitter can be considered as studying the overall sentiment of a tweet. There are two general categories of sentiment analysis methods in the Persian language, linked-base methods and, content-based methods. In this study, we implement a new link-based method for improving opinion classification in the Persian language. To compare with the content-based method, we implement a content-based method using Naive Bayes Method with two different weighting Methods: TF/IDF and Chi-Square. The TF/IDF method has good results in previous Persian language studies. The Chi-Square method has not been used in the Persian language researches, but the accuracy is fairly good in English. The results show that the improvement in the language-independent methods is remarkable and is in accordance with this research, the precision of the proposed algorithm for positive and negative comments was 98.87% and 97.87%, and the recall value for positive and negative comments was 99.24% and 96.84% respectively. The results also show that because of complexities in Persian syntax and lack of proper natural language processing tools in Persian, content-based algorithms operate poorly compared to English.

Indonesia with a population of 250 million is a large market, Millennials tend to be more adaptive to the development of communication technology [1]. There are lot of opportunities that are used by various groups, one of which is the need to use smartphones that can make it easier for people to exchange information [2].  The shift in sales of smartphone brands in Indonesia is influenced by massive advertising carried out by smartphone vendors (smartphone capitalists) to consumers [3]. The enthusiasm of the community in welcoming this platform is so great, lot of comment about smartphone brand stated by public is an interesting thing to be processed to be information. Utilization of that information requires analytical techniques so that the produced information can help many parties. The method used in this study is Naïve Bayes classification method which is a learning technique for data mining algorithms that uses probability and statistical methods [4]. This method is used to classify comments given by the community to smartphone brands. The comments given in this application will later be classified into positive, negative, and neutral comments. The purpose of this study was to find out how much positive, negative and neutral comments the community gave to smartphone brands, so that later it would facilitate the smartphone brand in providing policies or development in the future.

With the development of technology, consumers gather information in the internet environment before making a purchase decision in the tourism sector and the food and beverage sector under it. The positive and negative comments of consumers on the online review sites affect the decision making and buying behavior. The purpose of this study is to determine the negative customer comments on Tripadvisor regarding the restaurants operating in Sirince. Negative comments made for restaurant businesses were evaluated according to the criteria of food, staff, price, service, atmosphere and cleanliness and subjected to content analysis. It has been revealed that the criteria that restaurant customers make most negative comments are food, service and price. It has been determined that the complaints under these criteria arise due to bad and unpleasant food, cold service, slow service and expensive price. The study was ended with suggestions for restaurant managers and alternatives for future work to eliminate the problems.