Positive Sentiment Research Articles

Supervised Sentiment Analysis of COVID-19 Tweets: Insights into Public Reactions and Trends

In the modern digital era, social media platforms have experienced exponential growth, serving as a medium for individuals to share their emotions and opinions. Twitter, a prominent social media platform, hosts a substantial number of users who regularly express their thoughts through tweets. This research paper focuses on conducting sentiment analysis on tweets pertaining to the COVID-19 situation in India. By utilizing a lexicon-based approach using TextBlob, we determine sentiment scores for each tweet. These scores are then used to categorize the tweets into Positive, Negative, or Neutral sentiment categories. To further enhance the accuracy of sentiment classification, we employ the Naïve Bayes machine learning model, achieving an impressive accuracy rate of 87.65%. The sentiment analysis results obtained through this study provide valuable insights for governmental bodies and other organizations, enabling them to formulate effective policies in order to address similar issues in the future.

Development of the cognitive-behavioral communication (CBC) framework: A novel approach for mass communication in contact tracing during infectious disease outbreaks

The Cognitive-Behavioral Communication (CBC) Framework was developed to improve public compliance with contact tracing during infectious disease outbreaks, addressing the gaps exposed by the COVID-19 pandemic in the United States. These gaps, including public distrust, misinformation, and privacy concerns, hindered traditional contact tracing efforts. The CBC Framework integrates mass communication theories, cognitive-behavioral strategies, and AI-driven message personalization to enhance public engagement. The framework was tested through comparative field trials, with one group receiving personalized communication via the CBC model and another receiving standard public health messages. Results showed that the CBC group achieved a significantly higher compliance rate (85.6% vs. 66.4%) and a 33% reduction in privacy concerns. AI-driven messages also increased relevance over time, with 86.2% of participants reporting improved message relevance. Public trust was significantly higher in the CBC group, with positive sentiment at 67.5% compared to 45.2% in the control group. These findings have particular relevance to the U.S., where challenges like distrust and privacy concerns have impeded the effectiveness of contact tracing. The CBC Framework presents a scalable solution to improve compliance, trust, and overall public health outcomes during future outbreaks.

Tiktok as a Source of Education and Misinformation in Lichen Sclerosus.

The purpose of this study is to evaluate the content, delivery, and quality of medical information for vulvar lichen sclerosus on the social media platform TikTok. This is a descriptive, cross-sectional study. Using the third-party data scraping tool Apify, TikTok videos tagged with #lichensclerosus or "lichen sclerosus" were identified and sorted by view count. A sample of 100 videos was reviewed by 2 independent reviewers, excluding those not discussing lichen sclerosus. Videos were assessed using a coding document, the Patient Educational Materials Assessment Tool, and the DISCERN instrument. Interrater reliability was measured, and statistical analyses included Fleiss' kappa, intraclass correlation coefficient, t tests, and Wilcoxon rank sum test with Holm-Bonferroni correction. Content creators included patients (46%), health care professionals (30%), and others. Topics focused on clinical disease (52%) and treatment (48%). Evidence-based medicine was discussed in 71.7% of treatment-related videos, while 51.7% included nonevidence-based approaches, with a neutral or positive sentiment. Videos discussing topical steroids often had negative sentiments. Quality assessment revealed 61% of videos were understandable, 27% actionable, and 46% contained misinformation. Videos by health care professionals had less misinformation and higher quality scores compared to patient-generated content. Commercially biased videos were more understandable but contained more misinformation. TikTok serves as a significant platform for sharing information on lichen sclerosus, but nearly half of the content contains misinformation. Health care professionals need to engage in social media to provide accurate information and counteract misinformation. Enhanced collaboration with patient advocates and careful resource sharing can improve the quality and reliability of medical information available online.

Learners’ attitudes towards blended learning at Nguyen Tat Thanh University

In many educational settings accross the world, including Vietnam, blended learning has grown quickly due to evidence of its benefits over traditional classroom or online instruction alone. The study examined the attitudes of students at Nguyen Tat Thanh University toward the existing English blended learning course with the overall goal of enhancing the quality of language teaching and learning. A total of 367 students non-majoring in English language participated in the survey conducted at Nguyen Tat Thanh University in Ho Chi Minh City, Vietnam. The 44-item survey was used to collect the data, and SPSS version 20.0 was used for statistical analysis. The results demonstrated that most students had very positive sentiments towards four factors: the course's overall quality, the quality of the instructors, the course material, and the supportive services. However, a few students expressed dissatisfaction with the technology system, as they frequently encounter technical issues when engaging in online learning outside of the classroom.

Developing an mpox risk assessment matrix for social listening and infodemic insights generation

Abstract Issue Any effective response to an infodemic, requires prioritizing public health response on identified narrative themes that pose highest risk to health and well-being of populations. High-risk narratives often require more swift and far-reaching action and low-risk narratives require little or no response. Description of the problem Different countries are assigned different levels of risk to specific narratives or concerns, recognizing that for some issues there may be a large gap between the true public health threat and what is generating lots of conversations, concerns and press coverage that indicate a high degree of risk perception that may be unwarranted. WHO developed a risk matrix to apply to global analysis of questions, concerns and narratives about mpox, which resulted in biweekly infodemic intelligence reports that risk-assessed identified infodemic themes and corresponding recommendations for action. Results Risk assessment criteria were defined in the context of WHO’s global infodemic insights analysis for low, medium and high risk and positive sentiment in mpox global outbreak context before conducting the integrated analysis so that insights were reproducible and infodemic response actions were prioritized. This implemented the WHO/UNICEF training manual on 6 steps to produce an infodemic insights report, which includes tools on how to develop and apply a risk matrix for a specific public health question of concern and apply it in analysis. Lessons The approach resulted in rapid, systematic, reproducible and evidence-driven decision-making to respond to the infodemic accompanying the mpox global outbreak. Risk matrices must be adapted to the context of the public health event and the mandate of the health authority taking action. An mpox risk matrix for narratives included adaptation to reflect high-risk from stigmatizing narratives, as well as narratives circulating around or within LGBTQI+ communities. Key messages • A risk matrix must be developed and adapted to the context of each health emergency and level of health authority action. • Prioritization for rapid infodemic response based on a risk assessment is that it facilitates more effective response in a health emergency.

Chinese Public Attitudes and Opinions on Health Policies During Public Health Emergencies: Sentiment and Topic Analysis.

By the end of 2021, the new wave of COVID-19 sparked by the Omicron variant spread rapidly due to its highly contagious nature, affecting more than 170 countries worldwide. Nucleic acid testing became the gold standard for diagnosing novel coronavirus infections. As of July 2022, numerous cities and regions in China have implemented regular nucleic acid testing policies, which have had a significant impact on socioeconomics and people's lives. This policy has garnered widespread attention on social media platforms. This study took the newly issued regular nucleic acid testing policy during the COVID-19 pandemic as an example to explore the sentiment responses and fluctuations of netizens toward new policies during public health emergencies. It aimed to propose strategies for managing public opinion on the internet and provide recommendations for policy making and public opinion control. We collected blog posts related to nucleic acid testing on Weibo from April 1, 2022, to July 31, 2022. We used the topic modeling technique latent Dirichlet allocation (LDA) to identify the most common topics posted by users. We used Bidirectional Encoder Representations from Transformers (BERT) to calculate the sentiment score of each post. We used an autoregressive integrated moving average (ARIMA) model to examine the relationship between sentiment scores and changes over time. We compared the differences in sentiment scores across various topics, as well as the changes in sentiment before and after the announcement of the nucleic acid price reduction policy (May 22) and the lifting of the lockdown policy in Shanghai (June 1). We collected a total of 463,566 Weibo posts, with an average of 3799.72 (SD 1296.06) posts published daily. The LDA topic extraction identified 8 topics, with the most numerous being the Shanghai outbreak, nucleic acid testing price, and transportation. The average sentiment score of the posts was 0.64 (SD 0.31), indicating a predominance of positive sentiment. For all topics, posts with positive sentiment consistently outnumbered those with negative sentiment (χ27=24,844.4, P<.001). The sentiment scores of posts related to "nucleic acid testing price" decreased after May 22 compared with before (t120=3.882, P<.001). Similarly, the sentiment scores of posts related to the "Shanghai outbreak" decreased after June 1 compared with before (t120=11.943, P<.001). During public health emergencies, the topics of public concern were diverse. Public sentiment toward the regular nucleic acid testing policy was generally positive, but fluctuations occurred following the announcement of key policies. To understand the primary concerns of the public, the government needs to monitor social media posts by citizens. By promptly sharing information on media platforms and engaging in effective communication, the government can bridge the information gap between the public and government agencies, fostering a positive public opinion environment.

Public sentiment and ethical considerations of ChatGPT in higher education: Insights from data analytics of conversations on platform X

In today’s fast-paced digital world, generative AI, especially OpenAI’s ChatGPT, has become a game-changing technology with significant effects on education. This study examines public sentiment and discourse surrounding ChatGPT’s role in higher education, as reflected on social media platform X (formerly Twitter). Employing a mixed-methods approach, we conducted a thematic analysis using Leximancer and Voyant Tools and sentiment analysis with SentiStrength on a dataset of 18,763 tweets, subsequently narrowed to 5655 through cleaning and preprocessing. Our findings identified five primary themes: Authenticity, Integrity, Creativity, Productivity, and Research. The sentiment analysis revealed that 46.6% of the tweets expressed positive sentiment, 38.5% were neutral, and 14.8% were negative. The results highlight a general openness to integrating AI in educational contexts, tempered by concerns about academic integrity and ethical considerations. This study underscores the need for ongoing dialogue and ethical frameworks to responsibly navigate AI’s incorporation into education. The insights gained provide a foundation for future research and policy-making, aiming to enhance learning outcomes while safeguarding academic values. Limitations include the focus on English-language tweets, suggesting future research should encompass a broader linguistic and platform scope to capture diverse global perspectives.

Automated urban landscape design: an AI-driven model for emotion-based layout generation and appraisal

The evolution of a city is significantly shaped by the design of its urban landscape. The advancement of artificial intelligence has substantially increased convenience for individuals. This research proposes an urban landscape layout model powered by artificial intelligence that automatically generates urban landscape design based on deep learning (URDDL) with two dimensions: emotional tendency and urban landscape appraisal. The input image represents land use and surrounding road conditions, while the output image depicts the selection of the main entrance and the internal spatial function layout. The Pix2Pix model is trained to learn the internal function layout based on varying land use and road conditions. Additionally, a domain-specific dictionary is constructed using an existing semantic resource vocabulary, where positive and negative sentiment words are compared with their corresponding sentiment values, focusing on categories such as Stimulate, Sense, and Action. Experimental results indicate that the absolute average error of the URDDL model is 91.31%, with a maximum error of 96.87%. The degree of fit is highly appropriate for evaluating the emotional prediction of urban landscapes. The findings demonstrate that the URDDL model outperforms traditional design methods regarding generated results, suggesting its potential for future applications in automated landscape design.

Mapping perceived sentiments in university campuses with varied landscape metrics and climatic conditions

A sustainable university campus should accommodate students to experience positive emotions, which can be evoked by sustainable landscape with green and blue spaces (GBS). This effect is location-dependent because local vegetative type is climate-determinative, but evidence is not sufficient for sentiments of people experiencing campus landscapes. Forty-seven university campuses were selected along a latitudinal gradient in mainland China, and 100 subjects were chosen per campus (50 indoor and 50 outdoor). Photos of the subjects’ faces on Sina Weibo were collected. Facial expressions were assigned happy, sad, and neutral scores (n = 4,334). The average temperature (AveT) and blue space area (BlueA) showed negative relationships with latitude, thereby generating neutral emotion scores for subjects at indoor and outdoor locations. The ratio of green space area to host campus was the only landscape metric that depressed the presentation of happiness and enhanced sadness levels. Large water bodies should be built on campuses to induce calmness, and a high ratio of green spaces should be avoided. Mapping results show that campuses in eastern regions (Beijing and Liaoning) tend to elicit positive sentiments more frequently.

Advancing Electric Engineering Education through Immersive Virtual Reality: Deep Learning and Evolutionary Algorithms for Image Stitching and Rectification in Virtual Lab Environments

Virtual Reality (VR) technology has emerged as a transformative tool in education, offering immersive and interactive experiences that enhance learning outcomes. This paper delves into the application of image stitching and rectification techniques to create a VR lab environment, specifically tailored for electrical engineering education. The importance of VR technology in education is explored, highlighting its role in promoting active learning and providing experiential learning opportunities. The primary emphasis of this Paper lies in the smooth incorporation of image stitching algorithms for the creation of panoramic perspectives, along with the implementation of rectification techniques to correct irregular borders within the stitched images. By utilizing Convolutional Neural Networks (CNNs) and Genetic Algorithms (GAs), the proposed approach optimizes the rectification process, resulting in visually cohesive representations. Demonstrating the utilization of the VR lab across a range of situations, such as examining power transfer and creating control panels for water pumps in irrigation initiatives, the immersive setting enables students to delve into intricate systems. The performance of the proposed method was evaluated using various metrics, including mean squared error, peak signal to noise ratio (PSNR), structural similarity index (SSIM), and Fréchet inception distance (FID). the combination of deep learning algorithm specifically (CNN) and optimization algorithm specifically (Genetic algorithm (GA)) led to an increase in the accuracy of the rectified images where the average PSNR reached 23.98, SSIM was 0.8066, and FID was 18.72. Regarding the users’ opinion about the generated environment by stitching and rectifying images, participants demonstrated consistent positive sentiments, with mean scores ranging from 3.65 to 4.03, all above the scale midpoint, and moderate variability indicated by standard deviation values ranging from 1.070 to 1.251, suggesting general favorability with some variation in responses. This experience empowers the users to gain insights and cultivate essential problemsolving abilities at a heightened level. Collaborative learning is facilitated, enabling students to engage in collaborative projects regardless of their physical location. Through the synthesis of image processing techniques and VR technology, this research contributes to the enrichment of educational experiences and the advancement of electrical engineering education.

Academic Coverage of Online Activism of Disabled People: A Scoping Review

Disabled people need to be activists given the many problematic lived realities they face. However, they frequently encounter obstacles in traditional offline activism. Online activism could be a potential alternative. The objective of this scoping review is to examine the extent and nature of the coverage of disabled people in the academic literature that focuses on online activism. We searched the abstracts in Scopus, Web of Science, and the 70 databases in EBSCO-HOST for the presence of 57 terms linked to online activism or online tools or places for online activism, which generated 18,069 abstracts for qualitative analysis. Of the 18,069 abstracts, only 54 discussed online the activism by disabled people. Among these 54 relevant abstracts, only one contained the term “Global South”. No relevant abstracts were found that contained the terms “Metaverse” or “Democrac*” together with “activis*”. Only two relevant abstracts contained the phrase “digital citizen*”. Out of the 57 terms, 28 had no hits. The thematic analysis identified 24 themes: 6 themes in 30 abstracts had a positive sentiment, 7 themes in 30 abstracts had a negative sentiment, and 11 themes present in 23 abstracts had a neutral sentiment. There were three main themes: the positive role and use of online activism; the technical accessibility barriers to online activism; and the attitudinal accessibility problems arising from ableist judgments. The intersectionality of the disability identity with other marginalized identities and the issue of empowerment were rarely addressed, and ability judgment-based concepts beyond the term’s “ableism” and “ableist” were not used. The study underscores the necessity for further research given the few relevant abstracts found. The study also indicates that actions are needed on barriers to online activism and that examples for best practices exist that could be applied more often. Future studies should also incorporate a broader range of ability judgment-based concepts to enrich the analysis and to support the empowerment of disabled activists.

Tree hierarchical deep convolutional neural network optimized with sheep flock optimization algorithm for sentiment classification of Twitter data

ABSTRACT The increasing volume of online reviews and tweets poses significant challenges for sentiment classification because of the difficulty in obtaining annotated training data. This paper aims to enhance sentiment classification of Twitter data by developing a robust model that improves classification accuracy and computational efficiency. The proposed method named Tree Hierarchical Deep Convolutional Neural Network optimized with Sheep Flock Optimization Algorithm for Sentiment Classification of Twitter Data (SCTD-THDCNN-SFOA) utilizes the Stanford Sentiment Treebank dataset. The process begins with pre-processing steps including Tokenization, Stop words Elimination, Filtering, Hashtag Removal, and Multiword Grouping. The Gray Level Co-occurrence Matrix Window Adaptive Algorithm is employed to extract features, such as emoticon counts, punctuation counts, gazetteer word existence, n-grams, and part of speech tags. These features are selected using Entropy–Kurtosis-based Feature Selection approach. Finally, the Tree Hierarchical Deep Convolutional Neural Network enhanced by the Sheep Flock Optimization Algorithm is used to categorize the Twitter data as positive, negative, and neutral sentiments. The proposed SCTD-THDCNN-SFOA method demonstrates superior performance, achieving higher accuracy and lesser computation time than the existing models, respectively. The SCTD-THDCNN-SFOA framework significantly improves the accuracy and efficiency of sentiment classification for Twitter data.

Nested Sentiment Analysis for ESG Impact: Leveraging FinBERT to Predict Market Dynamics Based on Eco-Friendly and Non-Eco-Friendly Product Perceptions with Explainable AI

In the current era, the environmental component of ESG is recognized as a major driver due to the pressing challenges posed by climate change, population growth, global warming, and shifting weather patterns. The environment must be considered a critical factor, and as evidenced by existing research, it is regarded as the dominant component within ESG. In this study, the ESG score is derived primarily from the environmental score. The increasing importance of the environmental, social, and governance (ESG) factors in financial markets, along with the growing need for sentiment analysis in sustainability, has necessitated the development of advanced sentiment analysis techniques. A predictive model has been introduced utilizing a nested sentiment analysis framework, which classifies sentiments towards eco-friendly and non-eco-friendly products, as well as positive and negative sentiments, using FinBERT. The model has been optimized with the AdamW optimizer, L2 regularization, and dropout to assess how sentiments related to these product types influence ESG metrics. The “black-box” nature of the model has been addressed through the application of explainable AI (XAI) to enhance its interpretability. The model demonstrated an accuracy of 91.76% in predicting ESG scores and 99% in sentiment classification. The integration of XAI improves the transparency of the model’s predictions, making it a valuable tool for decision-making in making sustainable investments. This research is aligned with the United Nations’ Sustainable Development Goals (SDG 12 and SDG 13), contributing to the promotion of sustainable practices and fostering improved market dynamics.

Do Financial Community’s Tweet Sentiments Affect the Stock Market? Evidence from an Emerging Economy

In the era of social media and internet, Twitter (presently ‘X’) has been recognized as the prominent factor for financial markets by multiple corporates and experts. The motivation of present research hinges on burgeoning popularity of Twitter in the financial community for sharing information and impact of their hidden sentiments on financial markets. Research on tweet-based sentiments and their influence on financial markets in India is still in the infancy stage. To cover this gap, the present study empirically investigates the predictive impact of financial tweet sentiments on the Indian stock Index’s returns over a short period. The paper scraped 22,816 tweets and Nifty 50 index’s intraday hour end values for January 2023. Using Natural Processing Technique, the overall sentiments of the financial community are found as highly positive for each trading hour. Next, it presents a low positive correlation between positive sentiments, sentiment polarity, and hourly returns. Lagged negative sentiments reflect the better ability to impact current period returns than positive sentiments and polarity. Additionally, sentiment polarity and returns reflect bi-directional causation for the future period impact. Investors, stock market traders, and other practitioners can employ the proven concept of linkage between tweet sentiments and returns for constructing effective investment and trading strategies. JEL Classifications: G41, D53, G17, G02, G19

Exploring information demand in the maritime industry: a Google Trends approach

ABSTRACT The fundamental impetus for this research is the documented wide use of Google and the ensuing applications of Google Trends in behavioural economics. Expanding the existing literature, we focus on addressing the question of whether information demand from individual investors can reflect a freight rates sentiment. This is achieved by analysing Google queries referring to investment in maritime stocks and other related markets. These searches comprise essentially a blend of information and intuition. More specifically, we examine whether the most frequently searched term and topic referred to Baltic Dry Index (BDI) can serve as a reliable and valid positive or negative sentiment market proxies. We find a negative impact on freight rates, capital markets, and the number of available dry bulk fleet, providing evidence aligned with the risk-averse investor theory. We unveil bidirectional interactions between our sentiment index and the two out of three examined maritime markets, supporting the rejection of the efficient market hypothesis. Our study also verifies the similar effects of information demand and media-produced sentiment indexes on the corresponding market. We succeed in confirming sentiment’s predictability out-of-sample for BDI. Beyond the academic value, our findings suggest a manner of enhancing the accuracy of freight rates predictions for practitioners.

Too much of a good thing: Addressing the shape of relationship between positive media sentiment and IPO performance

The influence of mass media sentiment on the IPO performance of newly listed firms has received increasing research attention in management and entrepreneurship research. However, prior literature assumes the beneficial role of positive media sentiment in investors’ evaluation of the firm yet overlooks the potential downside of too much positive media sentiment. Based on two theories, dual processing theory and celebrity theory, we develop and test a curvilinear relationship between positive media sentiment and newly listed firms’ IPO performance. Further, we suggest that IPO market conditions, offering an important information cue, moderate this curvilinear relationship. Drawing a representative sample of newly listed U.S. firms in recent decades, we find support for the proposed inverted U-shape relationship and a moderating effect of market conditions.

Automatic detection of fake reviews at marketplaces using expert-based features and consumers’ reactions

The article presents the results of a practical study of the features of fake reviews that are described by marketers and other experts. Due to the abundance of fake reviews on marketplaces, consumer trust falls not only in the seller or platform, but in the genre itself. The paper presents the results of automatic classification of reviews from Russian marketplaces into potentially fake and honest ones using modelling of features that experts call labels of a fake review (presence of template words, exclamation marks, emoji, positive sentiment), and machine learning algorithms. To solve the problem, a corpus of 6 288 texts from the Russian marketplaces Wildberries and Megamarket has been collected. The target variable (predicted class) is the ratio of likes and dislikes given to the review by other buyers. The best result is demonstrated by the support vector machine algorithm in binary classification into reviews with a low and high ratings (without neutral ones). The classification model confirms that the formal features identified by experts as indicating fake reviews indeed have predictive potential. The quality of the model is reduced by the imbalance in classes and insufficient number of reviews with buyer reactions in our corpus, which leaves room for further work.

Challenges and Solutions in Scaling Operations Globally: A Guide for U.S. Food and BeverageExecutives

Abstract—This paper explores the challenges and solutions associated with scaling operations globally in the food and beverage sector from a U.S. executive’s perspective. It examines why expanding internationally makes sense for companies with successful domestic products, highlighting the advantages of U.S. innovation and lower costs of innovation compared to local markets abroad. The paper outlines a strategic roadmap for global expansion, including test markets, partnerships with local firms, and focusing on countries with positive sentiments toward U.S. food safety standards. A hypothetical case study illustrates the growth process, emphasizing the need for internal buy-in, patience, and a long-term view. Potential pitfalls and strategies for success are discussed, along with competitive analysis and the role of disruption in scaling products. Index Terms—Global Expansion, Food and Beverage Industry, Consumer Packaged Goods (CPG), International Operations, U.S. Innovation, Market Entry Strategy, Competitive Analysis, Disruption

Sentiment Analysis of User Reviews on BSI Mobile and Action Mobile Applications on the Google Play Store Using Multinomial Naive Bayes Algorithm

Mobile banking services are designed to facilitate customer transactions. Bank Syariah Indonesia (BSI) and Bank Aceh also provide these online services through their respective applications, BSI Mobile and Action Mobile. The mobile banking apps aim to simplify customer transactions, which can be conducted remotely via several features, from transfers, payments, and purchases to zakat payments, by simply connecting to the internet. Therefore, this research aims to classify the sentiment of user reviews for BSI Mobile and Action Mobile applications on Google Play Store to understand the users' experiences. The Multinomial Naïve Bayes algorithm is used in this study, where the algorithm analyzes and classifies the user reviews into positive and negative sentiment categories. The study involves several stages, such as text preprocessing, sentiment visualization, splitting the data into an 80:20 ratio for training and testing datasets, and training the model using the Multinomial Naïve Bayes algorithm. The results of this study show that the Multinomial Naïve Bayes algorithm performs well in analyzing user sentiment for BSI Mobile and Action Mobile, achieving an accuracy of 78.7%, precision of 76.5%, recall of 86.2%, and an F1-score of 80.6% for BSI Mobile, and an accuracy of 85.6%, precision of 75%, recall of 75%, and an F1-score of 75% for Action Mobile. Additionally, the sentiment classification results reveal that 52.8% of BSI Mobile user reviews are positive and 47.2% are negative, while for Action Mobile, 35.1% are positive and 64.9% are negative. For BSI Mobile, 21,497 reviews express a positive sentiment with dominant keywords such as "updated," "good," "balance," "transaction," and "thank." Meanwhile, for Action Mobile, 274 reviews express a negative sentiment with dominant keywords such as "transaction," "application," "network," "register," "please," and "update."

Treatment Recommendation using BERT Personalization

This research work develops a new framework that combines patient feedback with evidence-based best practices across disease states to improve drug recommendations. It employs BERT as its free-text processing engine to deal with sentiment judgment and classification. The functionality of the system, named `PharmaBERT`, includes acceptance of drug review data as a comprehensive input, drug categorization when dealing with a wide range of treatments and fine-tuning the BERT-based model for gaining positive or negative sentiment towards specific medications. PharmaBERT categorizes various drugs and fine-tunes the BERT structure to perceive lots of possible sentiments for specific medications. Consequently, PharmaBERT brings all its training and optimization capabilities together and through this, the system reaches a higher accuracy of up to 91% thus showcasing the potency of the model in capturing patient sentiments. While being a BERT spin-off, PharmaBERT utilizes its own set of experienced techniques to comprehend and sense the health-related text input given by the patient, doctor, or pharmacist. It uses transfer learning, that is, it learns from language representations to adapt quickly to the intricacies of drug reviewing. Through PharmaBERT, healthcare professionals may expand their diagnoses with insights from patient feedback to constitute more neutral decisions.