No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
... Additionally, opinion mining facilitates the extraction of useful information from user-generated content, improving customer happiness and responsiveness for businesses. [5], [6]. Real-time market trends, public concerns, and new themes are captured by topic modeling and trend detection techniques, which have a big impact on crisis response management, targeted marketing campaigns, strategic corporate decisions, and public health monitoring. ...
... Therefore, certain pretreatment methods that are specifically designed to account for the grammatical inconsistencies of social media are essential for precise downstream natural language processing tasks. Therefore, it is crucial to create strong preprocessing techniques to lessen these problems and raise the caliber of ensuing studies [5]. ...
... Social media key natural language processing (NLP) challenges cover a wide range of applications tailored to the particularities of this field. Identifying and categorizing user-generated content into groups like positive, negative, or neutral sentiment is the goal of sentiment analysis, one of the most studied NLP tasks for social media [5]. Businesses, legislators, and researchers may keep an eye on consumer happiness, public views, and reactions to goods, services, or events with the aid of social media sentiment analysis. ...
Natural Language Processing (NLP) is an essential element of computational linguistics and artificial intelligence, enabling fluid interactions between humans and machines. Social networking networks produce substantial volumes of user-generated text daily, offering both opportunities and challenges for NLP researchers. Social media discourse's informal, dynamic, and context-dependent characteristics necessitate specific NLP techniques for precise processing and analysis. This study thoroughly examines NLP applications in social media, including essential tasks such as sentiment analysis, topic modeling, misinformation detection, and hate speech identification. It examines the influence of machine learning and deep learning methodologies, particularly transformer models, on the advancement of NLP capabilities. This study also emphasizes the ethical issues related to NLP-driven social media apps, including data privacy, algorithmic bias, and the regulation of misinformation. The paper continues by discussing emerging research paths, highlighting the necessity for adaptable and ethical NLP solutions in the changing social media environment.
... Untuk menganalisis sentimen berita, dapat menggunakan analisa yang dikenal sebagai analisis sentimen. Analisis sentimen (disebut juga opinion mining) adalah ilmu yang menganalisis pendapat, sentimen, evaluasi, penilaian, sikap, dan emosi seseorang terhadap suatu entitas seperti, produk, layanan, organisasi, individu, isu, topik, dan atributnya [5]. Analisis sentimen merupakan hal yang kompleks, melibatkan 5 tahapan yang berbeda untuk menganalisis data sentimen yaitu data collection, text preparation, sentiment detection, sentiment classification¸ dan presentation of output. ...
... Analisis sentimen memiliki banyak nama. Sering kali analisis sentimen disebut sebagai subjectivity analysis, opinion mining, atau appraisal extraction [5]. Analisis sentimen memiliki beberapa pendekatan untuk menyelesaikan permasalahannya. ...
... Sentiment analysis deals explicitly with identifying the polarity of text datawhether the expressed opinion in the text is positive, negative, or neutral. Pang and Lee [14] and Liu [6] have extensively explored sentiment analysis, establishing foundational models that employ both supervised and unsupervised learning techniques. These methodologies have been pivotal in parsing complex user-generated data, such as product reviews and social media posts [15], [16]. ...
... Employing advanced NLP techniques, such as deep learning models that understand the linguistic context better than traditional models, could significantly improve the classification and analysis of the user feedback [13]. A longitudinal approach to analyzing user feedback could also provide dynamic insights into how user sentiments evolve in response to changes and updates in the app, helping Duolingo align its offerings more closely with user needs over time [6], [37]. Additionally, future research could incorporate multilingual feedback analysis using multilingual NLP techniques, such as mBERT or XLM-R, and explore the use of word embeddings or deep learning models like LSTM or Transformer to capture more complex contextual relationships within the text. ...
As digital education tools gained prominence, user feedback played a crucial role in refining and personalizing learning experiences. This study analyzed over 100,000 Google Play reviews of the Duolingo language-learning app, using text classification techniques to extract key insights into user sentiment and preferences. By employing natural language processing (NLP) methods, specifically logistic regression and Naive Bayes classifiers, the study categorized feedback into four primary themes: content, instruction, performance, and user interface and user experience (UI/UX). Logistic regression achieved an AUC score of 0.812, precision of 0.904, recall of 0.900, and F1-score of 0.894, while Naive Bayes achieved an AUC score of 0.806, precision of 0.904, recall of 0.900, and F1-score of 0.894. Both models demonstrated an accuracy rate of 90%. The results indicated that content was the most significant concern for users, comprising 74.2% of all reviews, followed by instructional feedback (14%) and performance issues (9.1%). This analysis provided valuable insights for developers aiming to enhance Duolingo’s user experience by addressing content quality, improving pedagogical approaches, resolving technical issues, and refining the user interface. The findings also contributed to the broader field of educational technology by demonstrating the application of machine learning techniques in understanding user feedback at scale.
... Sentiment analysis [1], a subfield of natural language processing (NLP), involves the extraction of subjective information such as opinions, emotions, and sentiments from text data [2]. This field has garnered significant attention due to its wide range of applications and inherent challenges [3]. ...
... Foundational work in this domain was introduced by [17], who tackled key challenges such as text classification and feature selection. Building on these techniques, [2] extended sentiment analysis to different granularity levels-document-level, sentence-level, and aspectlevel-proving essential in finance, where nuanced investor sentiment offers deeper insights into market movements. ...
... This review explores the impact of cryptocurrency news on the earnings of publicly traded companies, using Generative AI (GenAI) models with the BERT framework for sentiment analysis. Research by Pang and Lee (2008), Liu (2012), and Devlin et al. (2018) have advanced sentiment analysis in NLP, while studies by Hagenau et al. (2013) and Zhang et al. (2020) have demonstrated the potential of AI in enhancing financial forecasting and decision-making processes. ...
This study is, the authors believe, a groundbreaking investigation into the impact of cryptocurrency news on the earnings of publicly traded companies. Using advanced Generative AI (GenAI) models and the BERT framework for sentiment analysis, we integrated comprehensive data from the Financial Modeling Prep API. This enabled us to employ a rigorous event study methodology and advanced machine learning algorithms. Valuable insights were derived from the BERT model, shedding light on the reasons behind abnormal returns and facilitating a thorough analysis of material and immaterial impacts. The study’s findings highlight the significant influence of both positive and negative cryptocurrency news on cumulative abnormal returns (CAR), particularly among firms deeply involved in crypto activities. Notably, deliberate news, including official announcements, had a more pronounced impact than unintentional ones on market reactions. This innovative approach provides actionable insights into financial services, investment management, and corporate communication, offering a framework for improving predictive models, investment decisions, and risk management strategies.
... SA is a novel research field that aims to reveal people's sentiments about an entity using computer science (Medhat et al., 2014). It is also considered a field of study within the scope of text mining (Liu, 2012). In addition, it is an analysis comprised of a part of NLP. ...
Introduction: Debate on the future of Artificial Intelligence (AI) has recently been polarized. Positive, negative, and neutral differences of opinion about AI have led to the need for further inquiry into this issue. In particular, establishing AI’s future potential of use by identifying the feelings and opinions of countries about AI is deemed significant for developing nationwide and regional AI strategies. In this regard, this study aimed to determine the emotional states of Turks toward AI. Social media platform was accordingly exploited since it is an important data source to determine individuals’ feelings and opinions. Methodology: User comments on the posts published by Turkish national news channels on YouTube were examined through Sentiment Analysis (SA). In the dictionary-based SA method implemented, consumer/follower comments were classified as positive, neutral, and negative according to their polarity scores. Results: Analyses indicated that 697 (48.6%) of user comments were positive, 380 (26.5%) were negative, and 357 (24.9%) were neutral. It was concluded that Turkish society’s feelings toward AI were generally positive. Discussion and Conclusions: YouTube users' current emotional states, with or without knowledge of artificial intelligence, may differ in the future. It might thus be viewed as predictable that in the future, users who are more positioned in these processes will experience certain shifts in their sentiment states toward specific issues, from positive to negative, from negative to positive, or from neutral sentimental states to positive or negative.
... This review explores the impact of cryptocurrency news on the earnings of publicly traded companies, using Generative AI (GenAI) models with the BERT framework for sentiment analysis. Research by Pang and Lee (2008), Liu (2012), and Devlin et al. (2018) have advanced sentiment analysis in NLP, while studies by Hagenau et al. (2013) and Zhang et al. (2020) have demonstrated the potential of AI in enhancing financial forecasting and decision-making processes. ...
This study is, the authors believe, a groundbreaking investigation into the impact of cryptocurrency news on the earnings of publicly traded companies. Using advanced Generative AI (GenAI) models and the BERT framework for sentiment analysis, we integrated comprehensive data from the Financial Modeling Prep API. This enabled us to employ a rigorous event study methodology and advanced machine learning algorithms. Valuable insights were derived from the BERT model, shedding light on the reasons behind abnormal returns and facilitating a thorough analysis of material and immaterial impacts. The study's findings highlight the significant influence of both positive and negative cryptocurrency news on cumulative abnormal returns (CAR), particularly among firms deeply involved in crypto activities. Notably, deliberate news, including official announcements, had a more pronounced impact than unintentional ones on market reactions. This innovative approach provides actionable insights into financial services, investment management, and corporate communication, offering a framework for improving predictive models, investment decisions, and risk management strategies.
... Subjectivity expresses a person's feelings, views, or beliefs, as opposed to objectivity, which reveals factual information about the world (Liu, 2012). For followers of influencers demonstrating expertise in a given topic, informative videos containing more objective statements than subjective ones tend to be better received by consumers, resulting in higher purchase intentions (Sokolova & Kefi, 2020). ...
Influencer marketing has evolved significantly, becoming more sophisticated and integral to brand success. Addressing the limitations of previous studies, this article presents an empirically validated categorisation of influencers by simultaneously considering their personal characteristics and content attributes. Using data from over 11,000 YouTube videos, we propose an empirical categorisation of six influencer profiles – Expert, Motivator, Attractive, Productive, Perfectionist, and Middle-of-the-road – based on several personal characteristics such as trustworthiness, originality, expertise, and over 40 linguistic elements. Moreover, we demonstrate the association between these profiles and different metrics of digital consumer engagement. This research advances the literature on social media influencers, offering valuable insights for developing effective influencer marketing strategies.
... One of the subfields of NLP that extensively leverages deep learning is sentiment analysis. Sentiment analysis aims to automatically classify texts into predefined categories (e.g., positive, negative, or neutral) based on emotions conveyed in the text [3]. ...
... Особый интерес вызывает то, как в новостном контенте отражаются разного рода оценки ИИ -тревожные, оптимистические, нейтральноописательные и т. д. Для распознавания и систематизации таких оценочных элементов успешно применяется сентимент-анализ [Liu 2012;Pang, Lee 2008], представляющий собой совокупность алгоритмических методов классификации тональности или эмоциональной окраски текстов. ...
The authors analyzed the discourse strategies and topics involved in the media image of artificial intelligence in Telegram news channels (RIA Novosti, Mash, Live broadcast • News, Topor Live). A quantitative and qualitative sentiment analysis of Telegram posts made it possible to interpret the results from a discursive perspective, taking into account the socio-political and cultural context, as well as to analyze the impact of these representations on public perception and discussions about technological progress. The data were samples using a GitHub parser and manually verified. The discourse interpretation and sentiment analysis involved ranking sentences into thematic blocks that constructed the image of artificial intelligence. The linguistic analysis of news contexts demonstrated mostly positive and neutral modality. The key linguistic means included verbs of creative, intellectual, and service actions. Verbs of destruction and emotional-evaluative vocabulary were responsible for the negative sentiment. The texts revealed a strong metaphorization of artificial intelligence as help. Artificial intelligence was represented as a technology that can benefit humans, but is potentially dangerous and requires strict control. The assessment method remains debatable for multitopic texts, as well as the relationship between sentiment analysis and manual analysis of the discursive representation of a media image.
... Analisis sentimen merupakan proses mengekstraksi, memahami, dan mengklasifikasikan opini atau perasaan dari suatu teks. Teknik ini digunakan untuk menentukan apakah suatu teks bersifat positif atau negatif [11]. Dalam penelitian ini, analisis sentimen digunakan untuk mengetahui opini masyarakat di platform X (sebelumnya Twitter) terkait kasus korupsi di PT. ...
Kasus korupsi PT. Pertamina (Persero) yang mencuat pada awal tahun 2025 telah menarik perhatian luas masyarakat, khususnya di platform X. Penelitian ini bertujuan untuk menganalisis sentimen masyarakat terhadap kasus tersebut menggunakan algoritma Support Vector Machine (SVM) sebagai metode klasifikasi. Data diperoleh melalui teknik tweet harvest dengan kata kunci "korupsi pertamina" dalam rentang waktu 24 Februari 2025 hingga 31 Maret 2025, menghasilkan 280 data yang terdiri dari 7 sentimen positif dan 273 sentimen negatif. Tahapan penelitian mencakup pre-processing data, seperti pembersihan teks, tokenisasi, normalisasi, penghapusan stopword, dan stemming. Namun, karena dataset awal menunjukkan ketidakseimbangan kelas yang signifikan, dengan dominasi sentimen negatif, maka dari itu, digunakan metode Synthetic Minority Over-sampling Technique (SMOTE) untuk meningkatkan keseimbangan distribusi data. Hasil pengujian menunjukkan bahwa model SVM tanpa SMOTE memperoleh akurasi 89%, sedangkan setelah penerapan SMOTE, akurasi meningkat menjadi 96%. Hal ini membuktikan bahwa SMOTE mampu meningkatkan kinerja model dengan memperbaiki keseimbangan data. Penelitian ini berkontribusi dalam pengembangan analisis sentimen berbasis kecerdasan buatan, terutama dalam memahami persepsi publik terhadap isu sosial
Accurate gold price forecasting is essential for informed financial decision-making, as gold is sensitive to economic, political, and social factors. This study presents a hybrid framework for multivariate gold price prediction that integrates classical econometric modelling, traditional machine learning, modern deep learning methods, and their combinations. The framework incorporates financial, macroeconomic, and sentiment indicators, allowing it to capture complex temporal patterns and cross-variable relationships over time. Empirical validation on an eleven-year dataset (2014–2024) demonstrates the framework effectiveness across diverse market conditions. Results show that advanced supervised techniques outperform traditional econometric models under dynamic market environment. Key advantages of the framework include its ability to handle multiple data types, apply a structured variable selection process, employ diverse model families, and support model hybridisation and meta-modelling, providing practical guidance for investors, institutions, and policymakers.
Textual analysis is grounded in conceptual schemes of traditional qualitative and quantitative content analysis techniques that have led to the hybridization of methodological styles widely used across social scientific fields. This entry delivers an extensive review of the origins and evolution of text analysis within the domains of traditional content analysis. Emphasis is given to the conceptual schemas and operational structure of latent semantic analysis, and its capacity to detect topical clusters of large corpora. Further, this entry describes the operations of Entity–Aspect Sentiment Analysis which are designed to measure and assess sentiments/opinions within specific contextual domains of textual data. Then, this entry conceptualizes and elaborates on the potential of streamlining latent semantic and Entity–Aspect Sentiment Analysis complemented by Correspondence Analysis, generating an integrated operational scheme that would detect the topic structure, assess the contextual sentiment/opinion for each detected topic, test for statistical dependence of sentiments/opinions across topical domains, and graphically display conceptual maps of sentiments in topics space.
Experiential tourism, emerging from the broader concept of experiential travel, emphasizes meaningful engagement with local culture, history, and environments, often through structured experiences facilitated by tourism operators. This paper explores the evolution of experiential tourism within the context of the experience economy, highlighting how shifts in consumer preferences and advancements in technology have driven the demand for personalized, immersive experiences. The study also examines the implications of these trends for small and medium-sized enterprises (SMEs) and the rise of multinational corporations in the tourism sector. Additionally, the relationship between experiential tourism and sustainable development is discussed, particularly in light of the UNESCO 2030 Agenda, which underscores the importance of preserving cultural heritage and promoting responsible tourism practices. The integration of digital technologies, including virtual and augmented reality, is also considered for its role in enhancing the tourism experience and supporting sustainability goals.
E-governance has transformed public administration, particularly in labour welfare programs. However, challenges remain in optimizing service delivery and ensuring equitable access. Sentiment analysis, powered by machine learning (ML), provides real-time feedback and predictive insights to address these issues. This chapter explores how integrating sentiment analysis with ML enhances e-governance efficiency in labour welfare. By leveraging these technologies, governments can optimize resource allocation and tailor interventions based on public sentiment. It also discusses challenges like data privacy, system interoperability, and transparency. Case studies demonstrate how ML-driven sentiment analysis can create more responsive and inclusive welfare services.
Cultural appropriation is a vast and complex domain spanning sociology, psychology, marketing, law, communication, and cultural studies. Based on more than 31,000 Reddit posts, this research aims to use computational text-mining techniques to analyze the major latent themes related to cultural appropriation. The study also visualizes the diachronic variation in sentiments over time. Results show that Reddit online communities focus on various topics related to cultural appropriation, including racism, cultural diversity, self-awareness, religion, and fashion appropriation. Sentiment and emotion analysis results reveal that although trust tops the list of emotions, there seem to be also some negative emotions such as fear, anger, and sadness. Analysis of the cultural appropriation emojis used by the Reddit community reveals that the face with tears of joy emoji is the most frequently used, followed by the rolling on the floor laughing (ROFL), the person shrugging, the loudly crying face, and the female sign (). Given the global reach of social media platforms, our research offers a novel perspective on understanding this phenomenon on Reddit as it captures the complex relationships between Reddit users’ comments and specific aspects of cultural appropriation.
This study investigates the dual role of customer feedback in shaping hotel brand reputations, examining how it can both construct and deconstruct brand value in the digital age. Drawing from attribution theory and justice theory, we analyze 48,352 TripAdvisor reviews from 1,000 hotels to explore how internal attributions and perceived injustice in reviews influence value co-destruction. CorEx topic modeling and IBM Watson Tone Analyzer are utilized to explore review content and emotion tones. The results reveal internal, stable attributions and perceived distributive, procedural, and interactional injustice provoke value co-destructive reviewing behaviors. Five customer-based and seven company-based antecedents of value co-destruction are identified that can aid hospitality brands in leveraging reviews to improve policies, services, staff training, and communications. This study offers a novel perspective by revealing how the qualitative aspects of customer feedback, particularly emotional and linguistic tones, play a critical role in value co-destruction. These insights not only enrich the academic discourse on brand value dynamics but also provide practical implications for hotel managers in enhancing brand image, performance, and responsiveness to customer feedback.
The public debate on vaccines has been particularly relevant in Italy due to the introduction of childhood vaccination mandates and anti-COVID-19 vaccines. Our exploratory study focused on (1) identifying the media’s portrayals of childhood and adult vaccination, (2) highlighting the narratives used to portray individuals opposing vaccines and/or vaccine mandates, and (3) investigating the use of the term “No-Vax”. To these aims, we collected 2890 Facebook posts published by the Italian National Press Agency (ANSA) between 1 January 2016 and 31 December 2023, via the (Meta) CrowdTangle application. Data were analyzed using quantitative and qualitative techniques. Results show the presence of four main vaccine-related narratives in the pre-pandemic period (2016–2019)—i.e., vaccination as threatened by fake news, as a lifesaving practice, as a political matter, as a subgroup requirement—and three narratives during the pandemic and post-pandemic period (2020–2023)—depicting vaccinations as a long-awaited achievement, as a social requirement, and as a tool in need of confirmation. The results further show that the term ‘No-Vax’ has some negative connotations and is unable to represent the diversity of vaccine-critical positions. The media’s role in shaping public opinion suggests a need for more nuanced reporting that acknowledges the diversity of views and concerns regarding vaccination. Future research should explore how different media outlets frame vaccine hesitancy and the impact of these narratives on public health communication.
With the rapid increase in opinion-rich content shared across the internet, text sentiment analysis has emerged as a vital tool in both academic research and industrial applications. Sentiment analysis typically involves classifying a piece of text as expressing positive, negative, or neutral emotion. Traditional approaches to text classification often require extensive feature engineering and rely heavily on tokenization and embedding techniques, making them resource-intensive and less adaptive to context. To address these limitations, Long Short-Term Memory (LSTM) networks—an advanced form of Recurrent Neural Networks (RNNs)—have been adopted for their ability to capture long-range dependencies in textual data. This study proposes a sentiment classification model based solely on LSTM architecture to analyze short texts and effectively extract context-aware sentiment patterns. Unlike conventional models, LSTM-based frameworks can learn temporal word relationships without explicit syntactic parsing or handcrafted features. By leveraging the memory capabilities of LSTM, the proposed model enhances sentiment categorization accuracy while maintaining a relatively lightweight computational profile. Experimental evaluations demonstrate the effectiveness of LSTM in capturing contextual semantics, making it a suitable choice for real-time sentiment detection tasks in dynamic and user-generated content environments.
In today's digital age, an online reputation is a crucial asset for any business. Poorly managed responses to negative reviews on social media can lead to significant costs and damage. Sentiment analysis provides an effective way to monitor and analyze online opinions, particularly in real-time, allowing businesses to track public sentiment regarding their products and services. This project leverages sentiment analysis on Twitter to harness the power of real-time data, enabling businesses to assess and respond to customer feedback promptly. By using Long Short-Term Memory (LSTM) models, this approach offers advanced capabilities in analyzing tweet sentiments, providing a deeper understanding of consumer sentiment and enhancing social media monitoring.One key improvement of this project over existing tools is the focused collection of data exclusively from Twitter, reducing noise and minimizing the risk of false results caused by irrelevant data sources. By analyzing user interactions on social media, beyond basic metrics like likes, shares, and comments, sentiment analysis seeks to uncover the underlying emotions and motivations of consumers, providing valuable insights for brands, public figures, NGOs, governments, and educational institutions.Existing sentiment analysis tools typically require a background in data science and advanced technical knowledge. However, this project introduces a user-friendly interface, allowing non-experts to easily access and interpret sentiment analysis results. The interface will display product reviews along with their corresponding sentiments, providing a seamless experience for the user. Additionally, the project incorporates a phrase-level sentiment analysis feature, which analyzes user-input phrases and predicts the sentiment behind them, offering a more granular and precise understanding of social media content.
Drawing upon previous research, this paper creatively combines different methods to conduct text analysis of cancer diaries related to cancer treatment on Chinese social media Weibo, and explores the sentiment, adoption of cancer frames, and psychological aspects of language. Initially, sentiment analysis technology is employed to investigate the sentiment in cancer diaries. Subsequently, the cancer frames in the text data are encoded, and similarity analysis is conducted to depict the associations among different frames. Lastly, the psycholinguistic dictionary of Linguistic Inquiry and Word Count (LIWC) and Multiple Correspondence Analysis (MCA) are utilized to explore the multivariate associations among different sentiment orientations, frames, and psychological states. Our study aims to aid in understanding the concerns and communication of patient groups regarding the cancer treatment process, provide actionable insights for healthcare professionals and researchers to better understand patient experiences and improve cancer care communication.
This introductory chapter begins by exploring human language and intelligence through six levels of linguistics, followed by a concise history of Natural Language Processing (NLP), highlighting its key components and applications. It serves as the foundation for understanding NLP concepts and technologies discussed in subsequent chapters. Additionally, this chapter provides the conceptual groundwork for Workshop #1: Basics of Natural Language Toolkit (NLTK), presented in Chap. 11.
Background
Growing bodies of evidence suggest that cannabis may play a significant role in both oncological palliative care and as a direct anticarcinogenic agent, but classification as a Schedule I substance has complicated research into its therapeutic potential, leaving the state of research scattered and heterogeneous. This meta-analysis was conducted to determine the scientific consensus on medical cannabis’ viability in cancer treatment.
Objective
The aim of this meta-analysis was to systematically assess the existing literature on medical cannabis, focusing on its therapeutic potential, safety profiles, and role in cancer treatment.
Methods
This study synthesized data from over 10,000 peer-reviewed research papers, encompassing 39,767 data points related to cannabis and various health outcomes. Using sentiment analysis, the study identified correlations between cannabis use and supported, not supported, and unclear sentiments across multiple categories, including cancer dynamics, health metrics, and cancer treatments. A sensitivity analysis was conducted to validate the reliability of the findings.
Results
The meta-analysis revealed a significant consensus supporting the use of medical cannabis in the categories of health metrics, cancer treatments, and cancer dynamics. The aggregated correlation strength of cannabis across all cancer topics indicates that support for medical cannabis is 31.38× stronger than opposition to it. The analysis highlighted the anti-inflammatory potential of cannabis, its use in managing cancer-related symptoms such as pain, nausea, and appetite loss, and explored the consensus on its use as an anticarcinogenic agent.
Discussion
The findings indicate a strong and growing consensus within the scientific community regarding the therapeutic benefits of cannabis, particularly in the context of cancer. The consistent correlation strengths for cannabis as both a palliative adjunct and a potential anticarcinogenic agent redefine the consensus around cannabis as a medical intervention.
Conclusion
The consistency of positive sentiments across a wide range of studies suggests that cannabis should be re-evaluated within the medical community as a treatment option. The findings have implications for public health research, clinical practice, and discussions surrounding the legal status of medical cannabis. These results suggest a need for further research to explore the full therapeutic potential of cannabis and address knowledge gaps.
Social Media Analytics (SMA) is the practice of collecting, analyzing, and interpreting data from social media platforms to gain insights into user behavior, trends, and sentiment. As social media has become an integral part of daily life, the vast amount of data generated provides valuable information for businesses, marketers, and researchers. SMA encompasses various techniques, including text mining, sentiment analysis, trend detection, and user engagement metrics, to extract actionable intelligence. The primary goals of SMA are to understand audience preferences, track brand performance, enhance customer engagement, and predict emerging trends. With the help of machine learning, natural language processing, and data visualization tools, SMA transforms unstructured social media data into strategic business insights, helping organizations optimize their digital strategies, improve customer experiences, and make data-driven decisions. This paper explores the significance of social media analytics, the methodologies involved, and the impact of SMA on modern marketing and business strategies
L’article propose et défend un modèle de cartographie du raisonnement qui vise avant tout une description aussi exhaustive que possible des mouvements argumentatifs d’un texte en langue naturelle. Après avoir évoqué différentes impasses des approches existantes – la logique informelle, à l’origine de ces cartes, l’analyse computationnelle de l’argumentation et la linguistique du texte –, il est envisagé une approche fonctionnelle interdisciplinaire qui repense la manière de segmenter un texte et se confronte aussi à des données langagières pas toujours prises en compte dans l’approche philosophique de la carte de l’argumentation. Une typologie de fonctions argumentatives et un mode de représentation de ces fonctions est abordé et testé sur un exemple complexe. Une réflexion sur les enjeux des cartes, entre outil pédagogique et outil scientifique, traverse l’ensemble de l’article.
This article introduces and advocates a model designed for mapping reasoning, with a primary focus on exhaustively describing the argumentative moves within a text in natural language. The study begins by analyzing the limitations of existing methodologies, including informal logic – the foundational framework for these maps – computational argumentation analysis, and textual linguistics. It then proposes an interdisciplinary functional approach that revises text segmentation strategies and addresses linguistic elements often overlooked in philosophical discussions of argument mapping. The paper also presents a typology of argumentative functions and a methodology for depicting these functions, which are applied to a complex case study. Furthermore, the article explores the dual role of argument maps as both pedagogical and scientific tools, reflecting on their broader implications throughout.
Background
During the COVID-19 pandemic, public sentiment and demands have been prominently reflected on social media platforms like Weibo. Understanding these sentiments and demands is crucial for governments, health officials, and policymakers to make effective responses and adjustments.
Objective
The study aims to analyze public sentiment and identify key demands concerning COVID-19 policies and social issues using Weibo data, providing insights to improve China’s policies and legal systems in public health emergencies.
Methods
The study used Python tools to collect public opinion data from Weibo regarding policy adjustments, social issues, and livelihood concerns. A total of 50,249 valid comments on 100 blog posts were collected from December 2019 to October 2023 in China. The SnowNLP algorithm was employed for sentiment analysis, Latent Dirichlet Allocation was used for topic clustering, and sampling coding was applied to further explore public demands by condensing the comment texts.
Results
The study categorized 100 blog posts into 23 important topics, with average sentiment scores ranging from 0.24 to 0.66. These scores ranging from 0 to 1 reflect sentiment polarity, where lower values indicate more negative public sentiment. The topics of material safety and information security management had the lowest scores, at 0.24 and 0.34, respectively. The analysis further revealed that the 23 topics could be classified into 57 subtopics, and a total of 101 concepts were identified through coding. The study found that public demands fall into five key categories: transportation and travel security, epidemic protection and health security, law building and policy implementation, social services and public demand, and education demand.
Conclusions
The study underscores the complexity of public sentiment during the epidemic, with significant concerns about material safety and information security management. Public demands span basic survival needs to higher-order concerns such as education and legal protections. The findings suggest that policy-making processes must become more responsive, transparent, and equitable, incorporating real-time public feedback and ensuring comprehensive policies and legal systems are in place to address multifaceted public demands effectively.
Artificial Intelligence (AI) is transforming marketing and customer engagement by facilitating businesses to predict trends, enhance decision-making, and improve customer experiences through technologies like machine learning, predictive analytics, and natural language processing. Important applications include customer segmentation, sentiment analysis, and predictive analytics for forecasting purchasing behavior. Chatbots and virtual assistants have automated customer service and boosted sales. However, AI adoption faces challenges such as data privacy concerns, regulatory issues, algorithmic biases, and resource constraints, particularly for small businesses. Despite these hurdles, advancements in deep learning and generative AI are unlocking new opportunities for personalized customer engagement through augmented reality (AR), virtual reality (VR), and customized content. The paper also explores future opportunities, such as integrating AI with blockchain for secure data sharing and leveraging the Internet of Things (IoT) for real-time insights into consumer behavior. The findings of the study indicate that AI, when responsibly applied, not only enhances marketing efficiency and personalization but also facilitates innovative strategies that redefine customer engagement and market responsiveness. A comprehensive analysis is presented on the topics mentioned above.
In the era of rapid digital communication, vast amounts of textual data are generated daily, demanding efficient methods for latent content analysis to extract meaningful insights. Large Language Models (LLMs) offer potential for automating this process, yet comprehensive assessments comparing their performance to human annotators across multiple dimensions are lacking. This study evaluates the inter-rater reliability, consistency, and quality of seven state-of-the-art LLMs. These include variants of OpenAI’s GPT-4, Gemini, Llama-3.1-70B, and Mixtral 8 × 7B. Their performance is compared to human annotators in analyzing sentiment, political leaning, emotional intensity, and sarcasm detection. The study involved 33 human annotators and eight LLM variants assessing 100 curated textual items. This resulted in 3,300 human and 19,200 LLM annotations. LLM performance was also evaluated across three-time points to measure temporal consistency. The results reveal that both humans and most LLMs exhibit high inter-rater reliability in sentiment analysis and political leaning assessments, with LLMs demonstrating higher reliability than humans. In emotional intensity, LLMs displayed higher reliability compared to humans, though humans rated emotional intensity significantly higher. Both groups struggled with sarcasm detection, evidenced by low reliability. Most LLMs showed excellent temporal consistency across all dimensions, indicating stable performance over time. This research concludes that LLMs, especially GPT-4, can effectively replicate human analysis in sentiment and political leaning, although human expertise remains essential for emotional intensity interpretation. The findings demonstrate the potential of LLMs for consistent and high-quality performance in certain areas of latent content analysis.
Aspect-based sentiment analysis is a long-standing research interest in the field of opinion mining, and in recent years, researchers have gradually shifted their focus from simple ABSA subtasks to end-to-end multi-element ABSA tasks. However, the datasets currently used in the research are limited to individual elements of specific tasks, usually focusing on in-domain settings, ignoring implicit aspects and opinions, and with a small data scale. To address these issues, we propose a large-scale Multi-Element Multi-Domain dataset (MEMD) that covers the four elements across five domains, including nearly 20,000 review sentences and 30,000 quadruples annotated with both explicit and implicit aspects and opinions for ABSA research. Meanwhile, we conduct experiments on multiple ABSA subtasks under the open domain setting to verify the effectiveness of several generative and non-generative baselines, and the results show that open domain ABSA as well as mining implicit aspects and opinions remain ongoing challenges to be addressed.
ResearchGate has not been able to resolve any references for this publication.