Sentiment Of Tweets Research Articles

Taxation, health system endowment and institutional quality: ‘Social media’ perceptions across Europe

In this paper we analyse how health system endowment and the quality of institutions impact perceptions towards taxation. We conduct a sentiment analysis of Twitter users' tweets to determine whether the impact of the Covid-19 health emergency has modified the attitudes of the citizens towards taxation in the four largest European countries: France, Germany, Italy and Spain. We use a difference-in-differences estimation strategy, comparing the average sentiments of individual tweets regarding taxation in different European NUTS-2 regions, before and after the spread of the Covid-19 pandemic. Our results highlight that in regions characterised by higher health system endowment people adopted more positive attitudes towards taxation with respect to those living in regions with low levels of health system endowment over the period 2019-2020. In addition, we show how higher quality institutions led to more positive perceptions in relative and absolute terms, suggesting a greater predisposition for a more progressive tax system.

A survey and study impact of tweet sentiment analysis via transfer learning in low resource scenarios

A survey and study impact of tweet sentiment analysis via transfer learning in low resource scenarios

Seeking and Providing Social Support on Twitter for Trauma and Distress During the COVID-19 Pandemic: Content and Sentiment Analysis.

The COVID-19 pandemic can be recognized as a traumatic event that led to stressors, resulting in trauma or distress among the general population. Social support is vital in the management of these stressors, especially during a traumatic event, such as the COVID-19 pandemic. Because of the limited face-to-face interactions enforced by physical distancing regulations during the pandemic, people sought solace on social media platforms to connect with, and receive support from, one another. Hence, it is crucial to investigate the ways in which people seek and offer support on social media for mental health management. The research aimed to examine the types of social support (eg, emotional, informational, instrumental, and appraisal) sought and provided for trauma or distress on Twitter during the COVID-19 pandemic. In addition, this study aimed to gain insight into the difficulties and concerns of people during the pandemic by identifying the associations between terms representing the topics of interest related to trauma or distress and their corresponding sentiments. The study methods included content analysis to investigate the type of social support people sought for trauma or distress during the pandemic. Sentiment analysis was also performed to track the negative and positive sentiment tweets posted between January 1, 2020, and March 15, 2021. Association rule mining was used to uncover associations between terms and sentiments in tweets. In addition, the research used Kruskal-Wallis and Mann-Whitney U tests to determine whether the retweet count and like count varied based on the social support type. Most Twitter users who indicated trauma or distress sought emotional support. Regarding sentiment, Twitter users mostly posted negative sentiment tweets, particularly in January 2021. An intriguing observation was that wearing masks could trigger and exacerbate trauma or distress. The results revealed that people mostly sought and provided emotional support on Twitter regarding difficulties with wearing masks, mental health status, financial hardships, and treatment methods for trauma or distress. In addition, tweets regarding emotional support received the most endorsements from other users, highlighting the critical role of social support in fostering a sense of community and reducing the feelings of isolation during the pandemic. This study demonstrates the potential of social media as a platform to exchange social support during challenging times and to identify the specific concerns (eg, wearing masks and exacerbated symptoms) of individuals with self-reported trauma or distress. The findings provide insights into the types of support that were most beneficial for those struggling with trauma or distress during the pandemic and may inform policy makers and health organizations regarding better practices for pandemic response and special considerations for groups with a history of trauma or distress.

Festive or Failure? Differences in Sentiments of Consumer Tweets on Amazon Festive Flash Sale before and after COVID-19

Amazon’s Great Indian Sale is an annual mega shopping festival and provides an opportunity for retailers selling on e-commerce platforms to increase their sales, traffic as well as visibility. This study aims to understand consumer behaviour through the analysis of user-generated content (UGC) on social media concerning Amazon’s Great Indian Sale from 2015 to 2020. This time period was especially significant since it allowed for the analysis of differences caused by the pandemic. It was further divided into pre-COVID-19, during COVID-19 and post-COVID-19 period. The study also aims to establish a causal relationship between the feelings of engagement of the UGC on Twitter, with the strategies of marketing and promotion of the companies. Twitter’s API was used to extract the UGC. A two-step methodology was used to extract meaningful insights. First, Latent Dirichlet Allocation (LDA) was used to identify major topics related to the Flash sales organized during the festive season. Next, Sentiment Analysis was used to classify each tweet into Positive, Negative and Neutral Sentiments in order to understand user opinions and feelings towards the marketing campaign. It was found that both during the COVID period as well as the post COVID period, topics like Deals & Offers and Exclusive Promotions were positively perceived by consumers. Topics like Pre-event Excitement, Diwali Shopping Experience, Customer Support, and Money-saving Opportunities generated positive feelings during the pre-COVID times. During COVID times, positive sentiments were generated by topics like Delivery Workers and Attractive Deals for Gifting. On the other hand, topics like Insults and Noise generated negative sentiments in both the COVID and post COVID periods. The findings of this study are significant not only to improve and devise marketing strategies and social media content for E-commerce companies, but also to provide an understanding of the effectiveness of festive sales.

Exploring the public sentiment of local community on major infrastructure development: Evidence from media news and Twitter data

ABSTRACT This study analyzes public sentiment on news sites and Twitter concerning the development of the Trans-Sumatra Toll Road (JTTS). Applying a lexicon-based approach to calculate sentiment scores, this study found a predominance of positive sentiments in news articles and tweets regarding the JTTS development. Employing multinomial logit estimation to examine the determining factors, this study however found no specific relationship between JTTS segment completion and public sentiment on news sites or Twitter. Nonetheless, the association changed and became noticeable on Twitter after controlling the effects of socioeconomic characteristics on sentiment scores. The development status of the JTTS segment influenced public sentiment considerably. This study highlights the potential of social media to facilitate public engagement and inform government decision making. Furthermore, it offers a comprehensive analysis of public sentiment regarding toll road infrastructure development by incorporating multiple stakeholders, including government agencies, business actors, community organizations (via news articles), and local communities (via Twitter data), and highlights various interests, concerns, and conflicts surrounding toll road projects. The investigation of socioeconomic and demographic factors in relation to public sentiment further contributes to our understanding of the underlying influences and dynamics that shape opinions on toll-road infrastructure through sentiment analysis and quantitative methods.

Stakeholders' “voice” through social media: the case of corporate social responsibility in sport

PurposePrevious corporate social responsibility (CSR) research has mainly revolved around the “usual target” (that is, fans and consumers) that invest money, time and energy in supporting their teams in isolation while largely ignoring individual members of the public. Building on social exchange theory and social media analytics, the authors examine the social outcomes of CSR aggregated from individual members of society's perceived benefits (intangible and psychological).Design/methodology/approachRaw data were drawn from the CSR-focused Twitter accounts of six professional leagues (i.e. @nbacares, @nflplay60, @InspireChange, @thewnbpa, @Pr_nhl, @Mlsworks and @Mlbsocial). The authors collected historical data from each CSR-focused Twitter account (N = 136,076) from March 2010 to September 2022.FindingsAfter conducting sentiment analysis of public perceptions, the majority of tweets (53%) were neutral, 39% were positive and 8% were negative. All CSR-related accounts received more positive tweets about their initiatives than negative ones did. The most prevalent positive topics are supporting the community, education, youth wellness and health and inspiring the young generation. The most prevalent negative topics were related to fake, hypocrite, hate and social justice.Originality/valueThe study contributes to the CSR-sport literature by incorporating members of the general public into the stakeholder ecosystem and empirically examining their perceptions of sport organizations' CSR activities. Also, by drawing on the social exchange theory and the unique nature of social media, the authors highlight when and how the public expresses positive, neutral and negative perceptions over time. Finally, it joins a small but growing body of research that adopts the application of big data to sport management, and it measures the sentiment, frequency, distribution and topics of tweets, thereby determining positive and negative public perceptions.

Arabic ChatGPT Tweets Classification Using RoBERTa and BERT Ensemble Model

ChatGPT OpenAI, a large-language chatbot model, has gained a lot of attention due to its popularity and impressive performance in many natural language processing tasks. ChatGPT produces superior answers to a wide range of real-world human questions and generates human-like text. The new OpenAI ChatGPT technology may have some strengths and weaknesses at this early stage. Users have reported early opinions about the ChatGPT features, and their feedback is essential to recognize and fix its shortcomings and issues. This study uses the ChatGPT tweets Arabic dataset to automatically find user opinions and sentiments about ChatGPT technology. The dataset is preprocessed and labeled using the TextBlob Arabic Python library into positive, negative, and neutral tweets. Despite extensive works for the English language, languages like Arabic are less studied regarding tweet analysis. Existing literature about Arabic tweet sentiment analysis has mainly focused on machine learning and deep learning models. We collected a total of 27,780 unstructured tweets from Twitter using the Tweepy SNscrape Python library using various hash-tags such as # Chat-GPT, #OpenAI, #Chatbot, Chat-GPT3, and so on. To enhance the model’s performance and reduce computational complexity, unstructured tweets are converted into structured and normalized forms. Tweets contain missing values, URL and HTML tags, stop words, punctuation, diacritics, elongations, and numeric values that have no impact on the model performance; hence, these increase the computational cost. So, these steps are removed with the help of Python preprocessing libraries to enhance text quality and consistency. This study adopts Transformer-based models such as RoBERTa, XLNet, and DistilBERT that automatically classify the tweets. Additionally, a hybrid transformer-based model is proposed to obtain better results. The proposed hybrid model is developed by combining the hidden outputs of the RoBERTA and BERT models using a concatenation layer, then adding dense layers with “Relu” activation employed as a hidden layer to create non-linearity and a “softmax” activation function for multiclass classification. They differ from existing state-of-the-art models due to the enhanced capabilities of both models in text classification. Hybrid models combine the different models to make accurate predictions and reduce bias and enhanced the overall results, while state-of-the-art models are incapable of making accurate predictions. Experiments show that the proposed hybrid model achieves 96.02% accuracy, 100% precision on negative tweets, and 99% recall for neutral tweets. The performance of the proposed model is far better than existing state-of-the-art models.

A decision-making framework utilizing machine learning techniques, based on analysing the sentiment of tweets

Before making any decision, especially a financial one, people use social media to chat to, learn about, and obtain advice from others. How can social media opinions be relied on and utilised economically? Due to social media sites like Facebook and Twitter, where millions of individuals offer real-time thoughts on anything using slang and emoticons, data science researchers have access to more unstructured yet useful information. Social media includes Twitter. 326 million active social media users sent 500 million tweets daily in July 2018, indicating the sector's rapid expansion. Tweets can be viewed, edited, and sent. Knowing how to interpret tweets, which might be facts or views, is helpful. Such a study may predict the stock market, election, response event, news, and subjectivity scales. Tweets are analysed using Sentiment Analysis (SA). On the other hand, Sentiment analysis provides a broad overview of tweet polarity and is not beneficial for decision-making. In this research, we design a natural language processing (NLP) Twitter SA model to overcome the challenges of tweet sentiment recognition and build a hybrid system that can recognise and classify sentiment in Twitter's real-time reactions to any topic. We use a three-classifier machine learning method to analyse tweets from the Sentiment140 dataset (Logistic Regression, Bernoulli Naive Bayes, and SVM). Our findings are supported by the Term Frequency-Inverse Document Frequency analysis (TF-IDF). After that, accuracy and F1 Scores are used to evaluate these classifiers.

The Development of a Deep Learning Model for Predicting Stock Prices

The volatility and complexity of the stock market make it difficult to predict stock values accurately. The primary goal of this paper is to overcome some of these difficulties by training the data to anticipate stock prices based on sentiment analysis of tweets. Using natural language processing (NLP) technology, the tweet sentiments were categorized into (positive - neutral - negative). The stock price was predicted using deep learning algorithms (CNNs, RNNs, LSTMs, BiLSTMs). Among the algorithms, (BiLSTM) achieved the best results in terms of accuracy (94%) and the others (CNN=90%, RNN=91%, LSTM=92%). The paper also confirms that the average MSE and RMSE (MSE=0.03552, RMSE=0.1882064) for the BiLSTM algorithm are achieved (MSE=0.03552, RMSE=0.1882064). As a result, the obtained results were better than previous studies.

Sentimental Analysis Of COVID-19 Vaccination Tweets Using Machine Learning Techniques

Twitter Sentiment Analysis applies sentiment analysis to data in tweets on social media platforms to identify user sentiment. The scope of research in this field has expanded steadily in recent decades. The reason is that the format of tweets is complex and difficult to process. The extremely small Tweet format introduces a whole new set of problems: B. Use of Slang and Acronyms. The global community remains affected by the ongoing COVID-19 pandemic, with its profound implications for the health and overall welfare of individuals worldwide. Our worldview has changed as a result of the pandemic. A vaccination campaign should be carried out among the population to stop the outbreak of the pandemic. However, people are unsure about vaccination, which is cause for concern. In this study, machine learning techniques were employed to assess the sentiment of public Twitter tweets concerning COVID- 19 vaccination. The researchers utilized two distinct ML methods, namely Support Vector Machine (SVM) and Logistic Regression (LR), to analyze the Twitter data and classify the tweets as either positive or negative. Keywords:Classifier, Twitter, Sentiment Analysis.

SENTIMENT CLASSIFICATION OF TWEETS WITH EXPLICIT WORD NEGATIONS AND EMOJI USING DEEP LEARNING

The widespread use of social media platforms such as Twitter, Instagram, Facebook, and LinkedIn have had a huge impact on daily human interactions and decision-making. Owing to Twitter's widespread acceptance, users can express their opinions/sentiments on nearly any issue, ranging from public opinion, a product/service, to even a specific group of people. Sharing these opinions/sentiments results in a massive production of user content known as tweets, which can be assessed to generate new knowledge. Corporate insights, government policy formation, decision-making, and brand identity monitoring all benefit from analyzing the opinions/sentiments expressed in these tweets. Even though several techniques have been created to analyze user sentiments from tweets, social media engagements include negation words and emoji elements that, if not properly pre-processed, would result in misclassification. The majority of available pre-processing techniques rely on clean data and machine learning algorithms to annotate sentiment in unlabeled texts. In this study, we propose a text pre-processing approach that takes into consideration negation words and emoji characteristics in text data by translating these features into single contextual words in tweets to minimize context loss. The proposed preprocessor was evaluated on benchmark Twitter datasets using four deep learning algorithms: Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN), and Artificial Neural Network (ANN). The results showed that LSTM performed better than the approaches already discussed in the literature, with an accuracy of 96.36%, 88.41%, and 95.39%. The findings also suggest that pre-processing information like emoji and explicit word negations aids in the preservation of sentimental information. This appears to be the first study to classify sentiments in tweets while accounting for both explicit word negation conversion and emoji translation.

Trends in bushfire related tweets during the Australian ‘Black Summer’ of 2019/20

Social media is widely used in emergencies, but the nature of the communication is poorly understood. We employed unsupervised topic modelling and sentiment analysis to analyse more than 80,000 Twitter tweets posted by users in Australia over a six-month period before, during and after the severe bushfires in 2019–2020, dubbed the ‘Black Summer’. While tweets about bushfire updates dominated, politics, donations and support, impacts and public opinion were also prominent themes. Social impacts were important in the early phase of the fires (Sept & Oct 2019) with health impacts discussed when the fires were most intense and ecological impacts becoming important in the recovery phase. Twitter users also talked about emergency responses, mainly evacuation, as the fires were starting, showing that Twitter played an important role in communicating advice to leave early to avoid harm. Although the bushfires caused death, destruction and disruption, the sentiment of tweets was balanced − 40% of all tweets were positive, 36% negative and the remainder neutral (24%). Sentiments shifted little over time, but some topics were strongly associated with the expression of negative sentiments. The unusual severity of the fires was attributed to climate change in the early and recovery phases but misinformation about arson as a cause briefly diverted attention from climate change in the middle. Twitter was used to express anger about a lack of action by the Australian government to address climate change. Unexpectedly, Twitter users from the most affected areas were more likely to post positive tweets than those further away, particular during recovery, suggesting both resilience and gratitude for support provided to them. Analysis of trend data on Twitter using machine learning has the potential to identify, in real time, whether appropriate messages are reaching and being disseminated effectively, provide early warning of potentially harmful misinformation and, if undertaken repeatedly, provide a metric against which responses to fire can be measured.

Leveraging Multiyear, Geospatial Social Media Data for Health Policy Evaluations: Lessons From the Philadelphia Beverage Tax.

Public reactions to health policies are vital to understand policy sustainability and impact but have been elusively difficult to dynamically measure. The 2021 launch of the Twitter Academic Application Programming Interface (API), allowing for historical tweet analyses, represents a potentially powerful tool for complex, comprehensive policy analyses. Using the Philadelphia Beverage Tax (implemented January 2017) as a case study, this research extracted longitudinal and geographic changes in sentiments, and key influencers in policy-related conversations. The Twitter API was used to retrieve all publicly available tweets related to the Tax between 2016 and 2019. Twitter. Users who posted publicly available tweets related to the Philadelphia Beverage Tax (PBT). Tweet content, frequency, sentiment, and user-related information. Tweet content, authors, engagement, and location were analyzed in parallel to key PBT events. Published emotional lexicons were used for sentiment analyses. A total of 45 891 tweets were retrieved (1311 with geolocation data). Changes in the tweet volume and sentiment were strongly driven by Tax-related litigation. While anger and fear increased in the months prior to the policy's implementation, they progressively decreased after its implementation; trust displayed an inverse trend. The 50 tweeters with the highest positive engagement included media outlets (n = 24), displaying particularly high tweet volume/engagement, and public personalities (n = 10), displaying the greatest polarization in tweet sentiment. Most geo-located tweets, reflecting 321 unique locations, were from the Philadelphia region (55.2%). Sentiment and positive engagement varied, although concentrations of negative sentiments were observed in some Philadelphia suburbs. Findings highlighted how longitudinal Twitter data can be leveraged to deconstruct specific, dynamic insights on public policy reactions and information dissemination to inform better policy implementation and evaluation (eg, anticipating catalysts for both heightened public interest and geographic, sentiment changes in policy conversations). This study provides policymakers a blueprint to conduct similar cost and time efficient yet dynamic and multifaceted health policy evaluations.

Sentiment Analysis Before Presidential Election 2024 Using Naïve Bayes Classifier Based On Public Opinion In Twitter

This study aims to determine the performance of the Naïve Bayes Classifier algorithm and sentiment analysis tested on a dataset obtained from Twitter social media scrapping with the topic of 2024 presidential candidates. Three candidates frequently discussed in public spaces were used as keyword parameters in data mining: #anis, #ganjar, and #pilpres2024, resulting in 3021 tweets extracted from 12/1/2022 to 31/1/2023, which were successfully converted to ".csv" format documents. Public opinions extracted from the dataset were then pre-processed using the Python programming language, resulting in 2157 cleaned tweets. The data that passed the pre-processing stage was then labeled as positive or negative sentiment. Sentiment analysis was performed using the Naïve Bayes Classifier algorithm with three testing experiments using different training and testing data compositions in each experiment. The results of the study showed that the best Naïve Bayes model was obtained in the first experiment with a 10% testing data and 90% training data composition, resulting in 71% accuracy, 93% precision, 66% recall, and an f-measure score of 77%. The conclusion of the study is that the electability of the 2024 presidential candidates shapes public opinion and generates public sentiment in the form of positive and negative tweets. Positive tweets had a higher percentage of 71.5% (1543), while negative sentiment tweets accounted for 28.5% (614). Further research is expected to produce different information by using different classification algorithms and larger data sets.

The Role of Social Media In Promoting The Safety Of Women in Indian Cities

In every city, harassment and violence becomes one of the major problems for women. Further, women’s personal life is suffered by the bullying and abusive content presented in online social networking (OSN). Therefore, it is necessary to identify the women safety in OSN environment. However, the conventional methods failed to predict the maximum safety analysis. So, this work is focused on women safety prediction using decision tree (WSP-DT) classifier. Initially, twitter dataset is considered to implement the entire system, which is then pre-processed to remove the missing and unknown symbols. Then, natural language toolbox kit (NLTK) applied to perform the tokenization, conversion to lowercase, stop words identification, stemming and lemmatization of tweets. Then, text blob protocol is developed to identify sentiments of pre-processed tweets, which identifies the positive, negative and neutral polarities of tweets. Further, term frequency-inverse document frequency (TFIDF) is applied to extract the data features based on word and character frequency. Finally, decision tree classifier applied to identify the fake or genuine tweet based on multi-level training. The simulations conducted on twitter dataset show that the proposed WSP-DT classifier resulted in superior performance than the other methods.

Intelligence system for sentiment classification with deep topic embedding using N-gram based topic modeling

Multi-modal information outbreak is consistently increasing in social media. Classification of tweet sentiments using various information modalities will help the recommender systems to achieve success in digital marketing. Moreover, aspect-level sentiment analysis categorizes a target’s sentiment polarity in a specific environment. Using topic modelling in aspect-level sentiment analysis enables the identification of more accurate aspect-based tweet sentiments. The existing sentiment classification techniques used for the development of recommendation systems do not focus on the aspect-based approach modelled using deep learning classifier with temporal analysis on the social media data. Hence, this paper proposes an efficient sentiment classification model that highlights the impact of topic modelling-based word feature embedding for improvising the classification of Twitter sentiments and product reviews based on temporal reasoning and analysis for performing predictive analysis. For tweets context analysis, Latent Dirichlet Allocation based topic modelling is used in this work which generates the topics. For each topic, the sentiment is calculated separately and the topic guided feature expansion is done using Senti-wordnet. Moreover, an extended deep learning classification algorithm called Long Short-Term Memory (LSTM) with word embedding and temporal reasoning(LSTMWTR) is proposed in this paper for improving the classification accuracy. Finally, the labelled data are classified using the existing machine learning algorithms namely Naïve Bayes, Support Vector Machines and also using the deep learning models such as Convolution Neural Network(CNN),LSTM, Recurrent Neural Networks (RNN) and the transformer model namelyBi-directional Encoder Representation from Transformers (BERT),Convolution Bi-directional Recurrent Neural Network (CBRNN) and the proposed deep learning algorithm namelyLSTMWTR. These sentiment classification algorithms have been evaluated with word embedding for tweet sentiment classification and product review classification. The results obtained from this work show that the proposed LSTMWTR algorithm emerges as the highly accurate model for tweet sentiment and product review classification.

A Modified Boosted Ensemble Classifier on Location Based Social Networking

One of the research issues that researchers are interested in is unbalanced dataclassification techniques. Boosting approaches like Wang's Boosting and ModifiedBoosted SVM (MBSVM) have been demonstrated to be more effective for unbalanceddata. Our proposal The Modified Boosted Random Forest (MBRF) classifier is aRandom Forest classifier that uses the Boosting approach. The main motivation of thestudy is to analyze sentiment of geotagged tweets understanding the state of mind ofpeople at FIFA and Olympics datasets. Tree based model Random Forest algorithmusing boosting approach classifies the tweets to build a recommendation system withan idea of providing commercial suggestions to participants, recommending localplaces to visit or perform activities. MBRF employs various strategies: i) a distancebasedweight-update method based on K-Medoids ii) a sign-based classifier eliminationtechnique. We have equally partitioned the datasets as 70% of data allocated fortraining and the remaining 30% data as test data. Our imbalanced data ratio measured3.1666 and 4.6 for FIFA and Olympics datasets. We looked at accuracy, precision,recall and ROC curves for each event. The average AUC achieved by MBRF on FIFAdataset is 0.96 and Olympics is 0.97. A comparison of MBRF and Decision tree modelusing 'Entropy' proved MBRF better.

Impact of Negation and AnA-Words on Overall Sentiment Value of the Text Written in the Bosnian Language

In this manuscript, we present our efforts to develop an accurate sentiment analysis model for Bosnian-language tweets which incorporated three elements: negation cues, AnA-words (referring to maximizers, boosters, approximators, relative intensifiers, diminishers, and minimizers), and sentiment-labeled words from a lexicon. We used several machine-learning techniques, including SVM, Naive Bayes, RF, and CNN, with different input parameters, such as batch size, number of convolution layers, and type of convolution layers. In addition to these techniques, BOSentiment is used to provide an initial sentiment value for each tweet, which is then used as input for CNN. Our best-performing model, which combined BOSentiment and CNN with 256 filters and a size of 4×4, with a batch size of 10, achieved an accuracy of over 92%. Our results demonstrate the effectiveness of our approach in accurately classifying the sentiment of Bosnian tweets using machine-learning techniques, lexicons, and pre-trained models. This study makes a significant contribution to the field of sentiment analysis for under-researched languages such as Bosnian, and our approach could be extended to other languages and social media platforms to gain insight into public opinion.

Sentiment Analyzing from Tweet Data’s Using Bag of Words and Word2Vec

Twitter sentiment classification is an artificial approach for examining textual information and figuring out what people's publicly tweets from a variety of industries are experiencing or thinking. For instance, a large number of tweets containing hashtags are posted online every minute from one user to some other user in the commercial and politics fields. It can be challenging for scientists to correctly comprehend the context in which specific tweet terms are used, necessitating a challenge in determining what is actually a positive or negative comment from the vast database of twitter data. The system's authenticity is violated by this issue and user dependability may be significantly diminished. In this study, twitter data sent to interpret movies were classified using various classifier and feature methods. In this context, the IMDB database consisting of 50000 movie reviews was used. For the purpose of anticipating the sentimental tweets for categorization, a huge proportion of twitter data is analyzed. In the proposed method, bag of words and word2vec methods are given by combining them instead of giving them separately to the classifier. With both the suggested technique, the system's effectiveness is increased and the data that are empirically obtained from the real world situation may be distinguished well. With experimental efficiency of 90%, the suggested approach algorithms' output attempts to assess the reviews tweets as well as be able to recognize movie reviews.

FINE-GRAINED SENTIMENT ANALYSIS IN SOCIAL MEDIA USING GATED RECURRENT UNIT WITH SUPPORT VECTOR MACHINE

Social media platforms are widely used to share opinions, leading to a large growth of text data on the internet. This data can be a key source of up-to-date and inclusive information by conducting sentiment analysis. Typically, sentiment analysis research classifies binary based on the polar values generated. However, this has its limitations, such as classifying sentences containing positive and negative expressions, leading to incorrect predictions. Fine-grained sentiment analysis provides more precise results by associating values with more than two classification targets. The objective of this study is to carry out sentiment analysis at a fine-grained level related to public policy in Indonesia using the GRU-SVM model with feature extraction and expansion techniques. However, sentiment analysis research still faces challenges in NLP. Deep learning have successfully overcome the challenges of traditional machine learning models in terms of efficiency and performance. This study proposes GRU-SVM model. GRU is used because it can adaptively control dependencies, making it more efficient in memory usage, while SVM is used as it is state-of-the-art in sentiment analysis. Result of the study show that the selection of word representation techniques, the addition of feature extraction techniques, datasets, data ratios, and feature expansion are crucial in the model testing process. The GRU-SVM model achieved the best performance with an accuracy of 96.02%. Overall, the results of this study demonstrate that the GRU-SVM method is effective in analyzing sentiments in Indonesian tweets.