Publication Search

Rohingya in Indonesia has become trending conversation on social media. Sentiment analysis can get public responds. Big data makes the problem time efficiency labeling process, therefore the lexicon dictionary is needed for the labeling process. Data is growing and circulating very rapidly so it takes a fast and efficient time. Although it is fast and makes it easier to solve problems, it is still necessary to question the accuracy produced when using the lexicon labeling. A comparison of the labeling process between the InSet lexicon and the VADER lexicon was conducted to determine the accuracy of the labeling. It was done by combining lexicon with machine learning method of support vector machine and TF-IDF weighting and accuracy result calculated using confusion marix. Data from social media X as many as 9117 lines and labeled with InSet lexicon result 5241 negative sentiments, 1369 positive, and 521 neutral. Then the labeling results with VADER produced 2749 positive, 2523 negative, and 1881 neutral. After labeled, processed SVM and calculated accuracy with results of InSet lexicon accuracy having an average of 85.8% while the VADER SVM lexicon has an average of 82.65%.  

Research This investigate consumer sentiment analysis to halal products using social media data with utilise intelligence artificial intelligence (AI). With background behind increasing estimated market value of halal products reach USD 2.02 Trillion in 2024, understanding deep about opinion consumer become very important. Research This adopt approach quantitative, using secondary data from social media platforms such as Twitter, Instagram, and Facebook. Through Natural Language Processing techniques and algorithms learning machine, sentiment analysis is performed For identify pattern positive, negative and neutral in perception consumers. Research results show that 60% of the total 10,000 reviews had positive sentiment, with halal food products receiving the highest positive sentiment. Factors influencing consumer sentiment include product quality, price, and transparency of information. In addition, the study found that the use of AI in sentiment analysis provides advantages in efficiency and accuracy, and is able to capture nuances in consumer opinions that are not Possible done by manual analysis. From the analysis this, can concluded that the marketing strategy of halal products must focus on improving quality and providing clear information about halal certification. This study not only provides insight for halal industry players, but also enriches the literature related to AI, sentiment analysis, and sharia economics.

This research examines the role of social media platforms in fostering civic engagement and political participation in Latin America. Through qualitative interviews and sentiment analysis, the study reveals how digital spaces amplify marginalized voices and mobilize communities for social change. However, it also highlights risks such as misinformation and polarization. Recommendations include leveraging social media for transparent communication and digital literacy programs.

The advancement of Artificial Intelligence (AI) has brought about significant changes in various industries, including public relations (PR) practices in companies. This research aims to explore the implementation of AI in corporate PR activities in Indonesia. Using a case study approach with in-depth interviews with PR practitioners from three major companies, this research reveals how AI is being used to optimise PR functions. The findings show that AI is primarily used to accelerate media and sentiment analysis, facilitate social media content management, and enhance personalisation and automation in marketing communications. However, there are still limitations to the implementation of AI due to resource constraints and regulatory factors. This research contributes to a better understanding of AI adoption in corporate PR practices in Indonesia and its future development potential. By examining real-world cases, it provides valuable insights into the opportunities and challenges associated with using AI for strategic communication efforts in an emerging market context.

The Indonesian beauty industry has experienced remarkable growth in recent years, positioning the country as a lucrative market for businesses operating in the beauty tech sector. This rapid expansion has fueled fierce competition, prompting many manufacturers to adopt digital marketing campaign strategies to attract consumers and boost sales. Amidst this dynamic landscape, Somethinc, a leading beauty tech brand, has achieved remarkable success, multiplying its sales by 14 times compared to the previous year and securing the top spot as the best-selling brand on e-commerce platforms with total sales of Rp 53.2 billion. In light of this competitive environment, media monitoring plays a crucial role in understanding the positive sentiment surrounding Somethinc across social media and online news platforms. This study employs a qualitative descriptive method, utilizing a netnographic approach and leveraging media monitoring analytical tools to gather data through the Brand24 application. The primary objective of this research is to identify the dominant sources of positive sentiment for Somethinc across social media and news platforms. Comparative analysis of positive sentiment, mentions, reach, and product discussions on social media and news platforms regarding Somethinc reveals a notable advantage in social media mentions.

The reopening of the Tiktok Shop and collaboration with Tokopedia has caught the public's attention. That's proven by a post on Instagram about their collaboration that received various responses. This research aims to conduct sentiment analysis to ascertain whether the public approves of the two's partnership. The Naïve Bayes Classifier method was used to analyze 641 comment data from the period 11 December 2023 to 11 February 2024. The results show the composition of positive, negative, neutral sentiment and unclassified data, as well as accuracy, precision and recall. Of the 641 data recorded, there were 269 neutral sentiment data, 194 negative sentiment data, 176 positive sentiment data, and 437 unclassified data. The Naïve Bayes Classifier model with oversampling techniques obtained an accuracy of 83%, precision of 81%, and recall of 93%.

On August 14, 2023, Indonesia had approximately 228 million social media users, a number that is expected to continue growing to reach 267 million by 2028. Social media can be used to spread both positive and negative information, and one of the various negative effects is cyberbullying. Consequently, much research is conducted in the field of machine learning to develop sentiment analysis. One crucial step in sentiment analysis is word weighting. The two most common word weighting methods are TF-IDF and Word2Vec. These methods can be compared to determine which one produces better classification results, allowing cyberbullying sentiments on social media to be detected more accurately. Based on nine test scenarios, the final results showed that TF-IDF performed better than Word2Vec in this study, with an accuracy of 84%.    

In the growing digital era, social media, especially Instagram, has become the main platform for people to communicate and express themselves. One of the most influential accounts is President Joko Widodo's official account, @jokowi, which is often in the spotlight with thousands of comments covering a wide range of sentiments, both positive and negative. In the midst of his popularity, sentiment analysis is key to understanding the public's views on Jokowi's leadership. This study aims to analyze public sentiment towards President Joko Widodo (Jokowi) through comments posted on his official Instagram account (@jokowi). By utilizing the Naïve Bayes Classifier method, this study collected data from 1000 comments which were then processed through various stages of the methodology, including data collection, preprocessing, weighting, k-fold cross validation, and method implementation. Through the preprocessing stage involving cleansing, stopword removal, stemming, and tokenizing, the comments were prepared for further analysis. Test results using k-fold cross validation show that the model has an average accuracy of 80.3%. In addition, evaluation using confusion matrix showed an accuracy of 84.1%, with a precision of 85.5% and recall of 92.4%. These results show that the Naïve Bayes Classifier method performs well in classifying positive and negative sentiments in the comments.  

In the rapidly developing digital era, social media such as Twitter has become part of everyday life and facilitates the rapid dissemination of information, including information about criminals. This research aims to analyze public sentiment towards information about criminals spread on Twitter using the Naive Bayes algorithm. This algorithm was chosen because of its simplicity and effectiveness in text classification. Data was collected through a crawling process from Twitter, followed by a preprocessing stage to remove noise. The research results show that public sentiment towards information about criminals on Twitter is divided into three categories: positive, neutral and negative. After classification, it was found that neutral sentiment increased significantly to 63.4%, while positive and negative sentiment decreased to 10.5% and 26.1%. These findings indicate that people tend to be more careful in reacting to sensitive information. This research provides important insights for related parties in managing information about criminals on social media and can be a reference for developing further policies and strategies.

The remaining food waste in Indonesia reaches around 46.35 million tons, with economic losses reaching 23 million to 48 million tons per year. This condition has led to various campaigns to reduce food waste from people concerned about the problem of food waste. However, the increase in food waste campaigns has yet to be accompanied by a decrease in the volume of food waste in Indonesia. This research aims to determine public sentiment toward food waste campaigns on Instagram social media and determine the accuracy of the methods used in data classification. The method used is the Naïve Bayes Classifier method. The results obtained were from a total of 118 data regarding the food waste campaign; 79% data showed that the public had a positive sentiment, and 21% other data had a negative sentiment. The accuracy results of using sentiment analysis were 78.94%; this shows that the performance of the Naïve Bayes method in classifying data is quite good.

The rapid growth of social media platforms has generated massive volumes of unstructured textual data containing valuable information about public opinions and sentiments. Extracting meaningful insights from this data has become increasingly important for decision-making in various domains, including business, politics, and social analysis. This study aims to evaluate the effectiveness of deep learning techniques for sentiment analysis of social media data, focusing on Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and a hybrid CNN-LSTM model. A quantitative experimental approach is employed, where datasets are preprocessed through text cleaning, tokenization, and feature representation using word embeddings. The models are trained and evaluated using standard performance metrics, including accuracy, precision, recall, and F1-score. The results indicate that all models perform effectively in sentiment classification tasks, with the hybrid CNN-LSTM model achieving the highest performance due to its ability to capture both local textual features and long-term contextual dependencies. This demonstrates that combining CNN and LSTM architectures enhances classification accuracy compared to individual models. Furthermore, the findings confirm that deep learning approaches are more robust in handling the complexity and noisiness of social media data compared to traditional methods. This study contributes to the development of more adaptive and accurate sentiment analysis models and highlights the potential of hybrid deep learning architectures for real-world applications.

GoPay as one of the digital payment applications in Indonesia faces challenges in understanding user perceptions in the midst of fierce competition. This study aims to develop a user review sentiment analysis model by comparing two approaches to text representation, namely TF-IDF and BERT, as well as two machine learning algorithms, namely Random Forest and Logistic Regression. Review data is obtained from the Google Play Store and processed through pre-processing, feature extraction, and sentiment modeling. The results showed that the combination of BERT + Logistic Regression provided the best performance with an F1 Score of 0.86, showing the superiority of BERT in understanding the semantic context compared to TF-IDF. An important feature analysis identifies financial-related words such as "duitnyaapakah" and "kompensasi" as key issues. This research makes a practical contribution by helping app developers improve the user experience through prioritizing relevant features and solutions to key problems complained of.

Twitter is one of the media used by the Indonesian people to express their opinions regarding the 2024 Presidential Election. However, there is no scientific calculation that can determine the tone of public opinion regarding the 2024 presidential election. In this study, sentiment analysis was carried out on the tweets of the Indonesian people related to the 2024 Presidential Election (Pilpres 2024). The purpose of this study is to find out the opinions of Indonesian Twitter users regarding the 2024 Presidential Election using Natural Language Processing (NLP) Technology and Long Short Term Memory (LSTM) algorithms. NLP techniques are used to understand natural language and extract meaning from tweet copy, and LSTM is used to analyze the accuracy and accuracy of classification. The data used in this study was 1,004 tweets with the topic "Presidential Election", this data researchers obtained through the process of crawling using the tweet harvest library. In this study, 53.2% had positive emotions, 3.5% had neutral emotions, and 43.3% had negative emotions. 78% accuracy, 67% precision, and 67% recall.

The development of the automotive industry continues to grow rapidly and has modifications. The Alva One company is famous for producing electric motorbikes with one of the launches of its newest electric motorbike called Alva One XP. The popularity of Alva One brand products has encouraged the implementation of product monitoring or brand monitoring from news reports via the website. This research aims to see the level of positive sentiment, mentions and reach of the most influential websites. The research method was carried out by quantitative descriptive analysis and using a positivsm approach. Data collection is assisted by Brand24 application tools which focus on news website media. The research results illustrate the extent of positive sentiment received, the number of mentions and reach of the websites that have the most influence on reporting on the Alva One brand.

Obtaining high-quality, diverse, accurate datasets for sentiment analysis has always been a significant challenge. Traditional approaches include annotators, which may introduce bias to datasets and are also time-consuming and expensive. These types of datasets may also not represent the variety needed to train robust and generalizable sentiment analysis models. This study introduces a novel combination of techniques to approach the problem with a novel solution. The proposed system, SentiGEN includes the use of a transformer, T5, fine-tuned and optimized using an evolutionary algorithm to generate high-quality, diverse, accurate data for sentiment analysis. The generated data is validated using XLNet to ensure high sentiment accuracy. This combination of technologies has proven successful based on the results derived from evaluating multiple models. From complex transformers such as BERT to more straightforward approaches like KNN, those trained using synthetic data demonstrated superior performance compared to their counterparts trained on real data. This enhancement in predictive accuracy was observed when evaluated on benchmark datasets such as SST-2 and Yelp. SentiGEN can generate high-quality, diverse, accurate, realistic data for sentiment analysis and successfully increased the performance of models trained on synthetic data compared to the same model trained on real data.

Artificial Intelligence (AI) has been applied to many human endeavors, and epidemiology is no exception. The AI community has recently seen a renewed interest in applying AI methods and approaches to epidemiological problems. However, a number of challenges are impeding the growth of the field. This work reviews the uses and applications of AI in epidemiology from 1994 to 2023. The following themes were uncovered: epidemic outbreak tracking and surveillance, Geo-location and visualization of epidemics data, Tele-Health, vaccine resistance and hesitancy sentiment analysis, diagnosis, predicting and monitoring recovery and mortality, and decision support systems. Disease detection received the most interest during the time under review. Furthermore, the following AI approaches were found to be used in epidemiology: prediction, geographic information systems (GIS), knowledge representation, analytics, sentiment analysis, contagion analysis, warning systems, and classification. Finally, the work makes the following findings: the absence of benchmark datasets for epidemiological purposes, the need to develop ethical guidelines to regulate the development of AI for epidemiology as this is a major issue impeding it’s growth, a concerted and continuous collaboration between AI and Epidemiology experts to grow the field, the need to develop explainable and privacy retaining AI methods for more secured and human understandable AI solutions.

Background: Public opinion analysis has become increasingly important in the digital era, where social media platforms generate large-scale textual data reflecting public perceptions toward environmental policies. Advances in Natural language processing (NLP) and machine learning enable systematic sentiment classification to support data-driven decision-making. Objective: This study aims to evaluate the effectiveness of several sentiment classification models in analyzing Indonesian-language social media data related to environmental policies. Method: The research employed a text mining pipeline including data crawling, preprocessing (case folding, tokenization, stopword removal, and stemming), and vectorization using TF-IDF. Three classification models Logistic Regression, Support Vector Machine (SVM), and Long Short-Term Memory (LSTM) were trained and evaluated using accuracy and F1-score metrics. Results: Experimental findings indicate that LSTM achieved the highest performance with 91.7% accuracy and 91.2% F1-score, outperforming SVM (88.5%) and Logistic Regression (84.2%). Sentiment distribution analysis shows that public opinion is dominated by positive sentiment (47.5%), followed by neutral (32.0%) and negative (20.5%). Overall: The results demonstrate that deep learning-based models provide more robust contextual understanding and more reliable sentiment mapping for environmental policy analysis.

Web novels have gained popularity in recent years as a form of literature that is published and consumed online via specialized platforms. User generated comments and reviews play an important role in the web novel platform, providing valuable insight into reader sentiment and feedback. Manually analyzing sentiment from a large number ofcomments would be very time consuming, an efficient automated approach was required. This study uses the Latent Dirichlet Allocation (LDA) method to identify sentiment patterns (positive, negative, neutral) in user comments on web novels and analyze their distribution as a whole. LDA, originally designed for topic modeling, has proven effective in sentiment analysis, helping to group comments into relevant topics and uncover general sentiments related to each topic. This study aims to use the LDA method to identify sentiment patterns (positive, negative, or neutral) in user comments on web novels and analyze the distribution of sentiment as a whole. The results show the effectiveness of LDA in sentiment analysis, achieving quite good results, with 72% accuracy, 80% precision, 72% recall, and 65% F1 score.

Indonesia's golden opportunity to take part in a world-class soccer competition at the U-20 World Cup competition was wiped out, as FIFA gave the decision to revoke Indonesia's status as host of the U-20 World Cup. Indonesian netizens who felt disappointed expressed their opinions and trended on social media Twitter. This research focuses on sentiment analysis of tweets using a combination of FastText embeddings method for word vectorization and using LSTM type RNN algorithm for sentiment classification. The dataset used totals 9,645 data consisting of 4,141 positive data and 5,504 negative data taken from March 29, 2023 to April 05, 2023. The test results on the LSTM model provide the best performance with an accuracy value of 74.92%, precision 74.74%, recall 74.92%, and f1-score 74.78%. The conclusion of this research is that the majority of datasets have negative sentiments, which means that people are more likely to give negative opinions than to provide support to Indonesian football which is experiencing problems. It is hoped that with this conclusion in the future people will better control their opinions and provide positive opinions when Indonesia is experiencing problems.

Big data is a collection of data that has a large volume, so traditional data processing technology is unable to handle it well. Marketplace is a platform that most people often use to shop online. On this platform there is a comments column for writing reviews of products that have been purchased. Consumer reviews have an important role in understanding customer perceptions and sentiments towards the products being sold. Classification uses the Support Vector Machine method. The goal is to classify consumer reviews into positive or negative sentiment categories. The test uses data from more than 300 data samples with the assumption of independence between the features in the data. The results of big data analysis of consumer sentiment reviews of health masks on the marketplace used the support vector machine method with an accuracy value of 88%. The results of the analysis can be concluded that the dominant results of scraping reviews on health mask products lead more to positive reviews. The results on wordcloud of negative reviews provide insight to improve the quality of masks which are still lacking in terms of thinness, straps breaking easily, tears, holes, rubber quality and product packaging.