Sentiment Analysis Methods Research Articles

Exploring Sentiment Analysis in Indian Regional Languages: Methods, Challenges, and Future Directions

Sentiment Analysis, pivotal in natural language processing, extends its reach beyond English to Indian regional languages like Hindi, Marathi, Kannada, Konkani, Bengali, Khandeshi, and Urdu. This paper presents a comprehensive survey of 32 research papers in this domain, examining methodologies, datasets, and techniques while emphasizing the significance of sentiment analysis in diverse linguistic contexts for enhancing customer relationship management functionalities. It underscores the necessity for future research and highlights the efficacy of machine learning techniques. By elucidating on computational challenges and outlining various sentiment analysis methods, this paper serves as a critical resource for researchers and practitioners, fostering advancements in sentiment analysis tailored to regional linguistic nuances. KEYWORDS Bag Of Words, Hindi, Kannada, RNN, Konkani, Malayalam, Marathi, Maximum Entropy, Naive Bayes, Sentiment Analysis, SVM, TF-IDF, Urdu.

Does Pollyanna hypothesis hold true in death narratives? A sentiment analysis approach

Pollyanna hypothesis claims that human beings have a universal tendency to use positive words more frequently and broadly than negative words. The present study aims to test Pollyanna hypothesis in medical death narratives at both lexical and text levels by using sentiment analysis and emotion detection methods, and to qualitatively analyze the contextual use of emotion words to deepen the understanding of doctors' emotions. Sentiment analysis showed a strong token-based linguistic positivity and a weak type-based negativity bias at the lexical level, and a general positivity bias at the text level, despite the gender of the doctors. Emotion detection discovered three prominent emotions of “joy”, “sadness”, and “anger”, and a greater diversity of negative emotions in contrast to positive emotions in medical death narratives. Contextual analysis revealed that emotion words associated with joy were primarily observed in contexts related to doctors' actions and behaviors aiming to benefit others and promote social wellbeing. Emotion words associated with sadness and anger were chiefly employed to describe situations involving patients' death and doctors' attitudes towards death. The results confirm Pollyanna hypothesis at both token-based lexical level and text level and falsify the hypothesis at type-based lexical level. Possible explanations are explored by contextual analysis, and theoretical analysis from the perspectives of cognitive linguistics and social psychology. The findings are expected to enrich the understanding of Pollyanna hypothesis as well as the junior doctors' emotional responses to clinical deaths.

Analyzing Public Sentiment Regarding the Qatar 2023 World Cup Debate Using TF-IDF and K-Nearest Neighbor Weighting

This research aims to uncover the sentiment of Twitter users regarding the polemics surrounding the 2023 Qatar World Cup using a text-based sentiment analysis approach. The research methodology involves collecting data from Twitter posts, encompassing discussions, opinions, and responses related to the Qatar World Cup 2023. The TF-IDF weighting is applied to identify significant keywords in each post, while the K-Nearest Neighbor algorithm is employed to classify sentiments as positive, negative, or neutral. The findings reveal a comprehensive picture of how the public perceives the Qatar World Cup 2023 on the Twitter platform. The results not only cover positive and negative aspects of online discussions but also identify trends and patterns of sentiment that emerge during specific periods.The application of these methods provides valuable insights into understanding the dynamics of public opinion related to international sports events through the lens of social media. The results of the analysis demonstrate that a majority of Twitter users express positive sentiments towards the Qatar World Cup 2023, highlighting excitement and anticipation. However, some negative sentiments also arise, primarily related to controversies and concerns about the event. The research further identifies temporal variations in sentiment, reflecting changing public perceptions over time.This research contributes to the development of sentiment analysis methods by using a combination of TF-IDF weighting and the K-Nearest Neighbor algorithm to delve into Twitter users' perspectives. Consequently, the findings have practical applicability for further research and implementation in managing the social impact and public perception of major sporting events like the World Cup. .

The Spatial Distribution of Sexual Harassment Sentiment in Twitter

In a cultural landscape where discussions about sexual harassment are often considered taboo, this study leverages social media as a platform for the anonymous sharing of experiences related to sexual harassment in Asia. By focusing on Kuala Lumpur, we applied grid-based data mining to analyze the spatial distribution of tweets related to sexual harassment, revealing significant distribution patterns. Moreover, this study utilized two advanced sentiment analysis methods: the Natural Language Toolkit (NLTK) and Azure Machine Learning (AML), to evaluate the polarity of sentiments expressed in these tweets. The findings demonstrate a notable variance in sentiment analysis results between NLTK and AML, with NLTK classifying a majority of tweets as neutral (63.7%), while AML identified a predominant positive sentiment (70%). This discrepancy highlights the complexities of sentiment analysis and the importance of selecting appropriate tools for specific research contexts. In examining the spatial distribution, it becomes apparent that tweets collected through the identified keywords are dispersed within the boundaries of Kuala Lumpur. However, there is a notable concentration in specific areas, particularly evident in the hotspot encompassing Taman Desa and Pantai Dalam. Understanding the sentiment's location enables us to delve deeper into the pronounced cluster of tweets, examining potential correlations with surrounding social and moral issues. This insight empowers us to address the issue through various means, including the implementation of Crime Prevention through Environmental Design and fostering community-based initiatives.

Symmetrical Adoption Pattern of the Digital Sharing Economy

Listing spare homes as tourist accommodations on applications like Airbnb has boosted consumers’ adoption of the digital sharing economy (DSE). This research paper aims to develop a variable selection methodology for factors influencing consumers’ adoption intention of DSE applications like Airbnb and UBER. The symmetrical adoption pattern (SAP) will assist industry practitioners in designing an accurate investment pattern for the available resources. The research examines feedback from travellers regarding utilized services to develop SAP. The authors adopt NCapture as a data extraction tool and NVivo 12 as a data analysis tool to develop SAP as a variable selection methodology. Sentiment, thematic, and cluster analysis methods of qualitative analysis were employed to extract 19 distinct variables of SAP out of available data and adapt it into the six constructs of the unified theory of acceptance and use of technology (UTAUT2). By identifying the ideal variable for each construct with SAP, the performed study also aims to broaden the understanding of theories linked to the UTAUT2 model.

PENINGKATAN DONASI DENGAN STRATEGI PEMASARAN DIGITAL PEDULISEHAT.ID MENGGUNAKAN FRAMEWORK CRISP-DM

Pedulisehat.id is forum that offers services in the form of online, real-time, and transparent fundraising to collecting various donations from the public to help patients with chronic diseases and need funds for the treatment process. There are several problems that occur at Pedulisehat.id. Based on the data that has been collected, the researcher draws several problems that occur at Pedulisehat.id. First, the success of Advertising (a promo from a program) to increase public interest in donating to one of the existing programs at Pedulisehat.id. Second, Marketing Cost expenses in the Pedulisehat.id division are not and are not yet comparable to the donations that come in from the public. This research uses descriptive and sentiment analysis method to assist each campaigner in increasing respondents to the campaigns carried out by each user in terms of collecting and inviting the whole community to donate. The process of making this research design using 5 stages of CRISP-DM from 6 existing stages, there are: Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation. The results of this study are expected to provide the best campaign suggestions to users running on Pedulisehat.id and predictive analytics techniques, which from the results of analytics based on existing data will display predictions that will be achieved by Pedulisehat.id.Keywords — Real time, Penggalangan dana, CRISP-DM, Media social, Advertising

Research on online public opinion in the investigation of the “7–20” extraordinary rainstorm and flooding disaster in Zhengzhou, China

Sorting out the changing characteristics of online public opinion triggered by a series of events in the investigation and assessment of major natural disasters is of great practical significance for optimizing the work of disaster investigation and assessment, governing the ecology of online public opinion, and enhancing the effect of comprehensive disaster reduction. In this paper, we collected relevant comments from several official media accounts, such as People's Daily, and evaluated their emotional color using a sentiment analysis method based on the BERT fine-tuning model. Furthermore, keyword co-occurrence semantic network theme analysis is conducted for texts presenting negative emotional overtones to assess the changes in public opinion hotspots. The impact of the relevant online public opinion characteristics and the release of the disaster investigation report on them was assessed in the context of the investigation report itself. Based on sorting out the characteristics of online public opinion on several related topics, targeted public opinion governance initiatives and relevant suggestions for improving the disaster investigation system are proposed. This paper is of positive significance for studying disaster public opinion and improving the effectiveness of emergency management of rainstorms and flooding disasters.

Research on Barrage Text Mining and Emotional Orientation in Online Interactive Learning Videos: Based on the Case Study of "Luo Xiang's Criminal Law"

This study uses text mining and sentiment analysis methods to analyze the barrage text information of "Luo Xiang's Criminal Law" online teaching videos and sort out the main emotional tendencies of users. The research found that user emotions change with the dispersion of hot topics, showing a decentralized characteristic, and at the same time, users still have a certain lack of capture for the content of online teaching videos in the interactive learning process. It is urgent to clarify the specific impact of technology on users. This study is helpful for deepening research on online interactive learning, clarifying user emotional tendencies, and providing a decision-making basis for accurately improving interactive learning effects.

An Effective Strategy for Sentiment Analysis Based on Complex-Valued Embedding and Quantum Long Short-Term Memory Neural Network

Sentiment analysis aims to study, analyse and identify the sentiment polarity contained in subjective documents. In the realm of natural language processing (NLP), the study of sentiment analysis and its subtask research is a hot topic, which has very important significance. The existing sentiment analysis methods based on sentiment lexicon and machine learning take into account contextual semantic information, but these methods still lack the ability to utilize context information, so they cannot effectively encode context information. Inspired by the concept of density matrix in quantum mechanics, we propose a sentiment analysis method, named Complex-valued Quantum-enhanced Long Short-term Memory Neural Network (CQLSTM). It leverages complex-valued embedding to incorporate more semantic information and utilizes the Complex-valued Quantum-enhanced Long Short-term Memory Neural Network for feature extraction. Specifically, a complex-valued neural network based on density matrix is used to capture interactions between words (i.e., the correlation between words). Additionally, the Complex-valued Quantum-enhanced Long Short-term Memory Neural Network, which is inspired by the quantum measurement theory and quantum long short-term memory neural network, is developed to learn interactions between sentences (i.e., contextual semantic information). This approach effectively encodes semantic dependencies, which reflects the dispersion of words in the embedded space of sentences and comprehensively captures interactive information and long-term dependencies among the emotional features between words. Comparative experiments were performed on four sentiment analysis datasets using five traditional models, showcasing the effectiveness of the CQLSTM model.

Application of Sentiment Analysis in the Reshot Method to Improve User Experience of the Hijra Bank Application

The first stage in simplifying User Experience (UX) using the RESHOT method is Refine the Challenge. This stage is carried out so that design practitioners know the problems or needs of application users through product research. However, the product research methods that can be carried out at this stage are very diverse and require time. This research aims to explain a product research method that can be used by design practitioners easily and quickly, namely the sentiment analysis method with the Naïve Bayes algorithm. Naive Bayes is a classification method based on simple probability. The results of this analysis will be used as a reference for improving UX using the RESHOT method on the Hijra Bank application. The results of product research using sentiment analysis obtained 2711 opinions originating from tweets on Twitter and user reviews of the Hijra Bank application on Google Play. Of the 2711 opinions, 149 had negative sentiment, with the most frequently mentioned opinions being "Customer" and "Data" with an analysis accuracy of 92%. The results of this analysis are converted into a hypothesis, which will later become a reference in designing interfaces using the RESHOT method.

Sentiment analysis of video danmakus based on MIBE-RoBERTa-FF-BiLSTM

Danmakus are user-generated comments that overlay on videos, enabling real-time interactions between viewers and video content. The emotional orientation of danmakus can reflect the attitudes and opinions of viewers on video segments, which can help video platforms optimize video content recommendation and evaluate users’ abnormal emotion levels. Aiming at the problems of low transferability of traditional sentiment analysis methods in the danmaku domain, low accuracy of danmaku text segmentation, poor consistency of sentiment annotation, and insufficient semantic feature extraction, this paper proposes a video danmaku sentiment analysis method based on MIBE-RoBERTa-FF-BiLSTM. This paper constructs a “Bilibili Must-Watch List and Top Video Danmaku Sentiment Dataset” by ourselves, covering 10,000 positive and negative sentiment danmaku texts of 18 themes. A new word recognition algorithm based on mutual information (MI) and branch entropy (BE) is used to discover 2610 irregular network popular new words from trigrams to heptagrams in the dataset, forming a domain lexicon. The Maslow’s hierarchy of needs theory is applied to guide the consistent sentiment annotation. The domain lexicon is integrated into the feature fusion layer of the RoBERTa-FF-BiLSTM model to fully learn the semantic features of word information, character information, and context information of danmaku texts and perform sentiment classification. Comparative experiments on the dataset show that the model proposed in this paper has the best comprehensive performance among the mainstream models for video danmaku text sentiment classification, with an F1 value of 94.06%, and its accuracy and robustness are also better than other models. The limitations of this paper are that the construction of the domain lexicon still requires manual participation and review, the semantic information of danmaku video content and the positive case preference are ignored.

Understanding Public Sentiment Towards Crypto Assets

This study aims to examine the evolution of public sentiment towards crypto assets. The research approach involves sentiment analysis methodology utilizing the SentiStrength v2.3 software. The data utilized in this study is secondary and obtained from scholarly publications indexed in the Scopus database. A total of 112 scholarly publications related to the research topic were identified. The research findings indicate that positive sentiment accounts for 27%, negative sentiment for 34%, and neutral sentiment contributes 39%. Further analysis reveals that the year 2023 marked a period with the highest number of sentiments, with positive sentiment dominating with 23 sentiments. This sentiment analysis method provides an in-depth understanding of the dynamics of public perception towards crypto assets, and these findings can offer valuable insights into the trends and shifts in sentiment that occur over time. Sentiment analysis can offer deep insights into how cryptocurrencies are perceived by the public and how these perceptions evolve in line with changes in social and economic contexts.

Chinese young adults' perception of fertility: Emotional contagion in fertility policy discussions on Sina Weibo

We investigated the reactions and preferences of the Chinese populace, particularly young adults, toward China's three-child policy, a recent strategy aimed at addressing the nation's aging population that allows couples to have up to three children. By monitoring Sina Weibo discussions related to fertility policies, we gauged public sentiment and potential emotional contagion. We collected 711 main posts and 26,515 responses and used the sentiment analysis method to identify the emotional tendencies of each post, thus determining if emotional contagion occurred. Results showed there were 631 posts with negative emotional contagion effects and 554 posts with positive emotional contagion effects. Further textual analysis indicated that the Chinese public maintains a pessimistic outlook on the three-child policy and expresses concern about the detrimental impact of social circumstances and the difficulties women encounter concerning fertility.

A CONSUMER SENTIMENT ANALYSIS METHOD BASED ON EEG SIGNALS AND A RESNEXT MODEL

E-commerce is becoming increasingly dependent on technologies such as consumer sentiment research at an ever-increasing rate. Its purpose is to recognize and comprehend the feelings and dispositions of customers by analyzing customer language and behavior as expressed in social media, online reviews, and other forms of digital communication. The proliferation of digital technology has resulted in an increase in the number and variety of channels through which customers can communicate their feelings. Gaining a comprehensive understanding of consumer sentiment may be of great use to businesses, as it enables them to better satisfy customer demands, enhance their products and services, improve their brand reputation, and ultimately increase their level of competitiveness. As a result, consumer sentiment research has evolved into a tool that is essential for the decision-making process in e-commerce as well as the management of customer relationships. Within the scope of this discussion, this study uses deep learning models to improve consumer sentiment research precision. The following is the list of the primary contributions that this paper makes. (1) Advancing the use of EEG signals as a basis for a method for analyzing customer feelings. This technique measures brain activity directly, thus avoiding the restrictions and ambiguities that come with relying on verbal expression. (2) The purpose of this study is to improve the overall performance of the model for analyzing sentiment by incorporating an attention mechanism into the ResNeXt model. This attention mechanism is intended to augment the model’s capacity to extract subtle characteristics. (3) The results of the experiments show that the strategy described in this study is effective in improving EEG-based sentiment analysis performance. When compared to standard text-based sentiment analysis approaches, this sentiment analysis model demonstrates greater objectivity, real-time capabilities, and multidimensionality when applied to consumer sentiment analysis in e-commerce.

Aspect-Level sentiment analysis based on fusion graph double convolutional neural networks

ABSTRACT Aspect-based sentiment analysis (ABSA) aims to analyze the emotional color contained in sentences or documents in more detail by classifying and evaluating different aspects and emotions in the text. However, the current research methods cannot effectively analyze the relationship between aspect words and context and extract grammatical information about sentences. Additionally, the extracted syntactic information is insufficient, and the combination of syntactic and semantic information is inefficient, leaving the model unable to correctly determine aspects' emotional orientations. This paper proposes an aspect-level sentiment analysis based on Fusion Graph Double Convolutional Neural Networks (FGD-GCN) to address these issues. Firstly, FGD-GCN proposes a multi-feature extraction module. Using BERT and bidirectional long-short-term memory models, this module extracts the hidden context between words. In addition, the positional attention module is used to capture important features in sentences, reducing noise and bias. Then, a semantic enhancement module is proposed, which fuses attention-focused information and feature information extracted from graphs to emphasize aspect words and context, and uses CNN model to classify on feature vectors. According to experiment results on three benchmark datasets, the model outperforms previous GCN methods for context-based aspect-level sentiment analysis.

Otel Yorumlarının Sınıflandırılmasında Makine Öğrenmesi Algoritmalarının Karşılaştırılması: Tripavdisor Yorumlarının Duygu Analizi

Sentiment analysis can help extract meaningful information from these data piles from various websites and social media and measure consumers' reactions by classifying consumers' emotions as positive, negative or neutral. The success of sentiment analysis varies according to feature selection, vector space selection and machine learning method. For this reason, determining the most successful method in sentiment analysis is still controversial and important. A limited number of studies have been conducted comparing the success of various machine learning methods in sentiment analysis of hotel reviews in English. Considering this gap, the purpose of this research is to determine the most successful machine learning algorithm for sentiment analysis of hotel reviews. For this purpose, 708 reviews for 5-star hotels in Istanbul were collected manually. Obtained data were classified as positive and negative using logistic regression, k-nearest neighbor, naive Bayes and support vector machine methods. Analysis results show that the logistic regression method was the most successful classification algorithm, with an accuracy rate of 0.92. It is followed by support vector machine (0.90), naive Bayes method (0.77) and k-nearest neighbor algorithms (0.66).

Rule-enriched Decision Tree Classifier for Conditional Sentence Sentiment Analysis

Conditional sentences are often used when people have to choose with some requirements. Conditional sentences account for more than 8% of user opinions. Although accounting for a considerable amount, for sentiment analysis methods, the sentiment expressed in conditional sentences is still analyzed as typical narrative sentences. This causes the approaches to fail to achieve maximum performance. To solve this problem, although some studies have proposed separate approaches to sentiment extraction and analysis for conditional sentences, there are very few, and the performance still needs to improve. This study proposes a new classifier based on a decision tree classifier model enriched with rules (called Rule-enriched Decision Tree Classifier (ReDTC)) to extract and analyze sentiments expressed in conditional sentences. ReDTC has been experimented on a dataset collected from English teaching websites. The performance gain demonstrates that the proposed ReDTC method significantly improved the performance in sentiment extraction and analysis in conditional sentences.

HYBRID RECOMMENDER FOR VIRTUAL ART COMPOSITIONS WITH VIDEO SENTIMENTS ANALYSIS

Topicality. Recent studies confirm the growing trend to implement emotional feedback and sentiment analysis to improve the performance of recommender systems. In this way, a deeper personalization and current emotional relevance of the user experience is ensured. The subject of study in the article is a hybrid recommender system with a component of video sentiment analysis. The purpose of the article is to investigate the possibilities of improving the effectiveness of the results of the hybrid recommender system of virtual art compositions by implementing a component of video sentiment analysis. Used methods: matrix factorization methods, collaborative filtering method, content-based method, knowledge-based method, video sentiment analysis method. The following results were obtained. A new model has been created that combines a hybrid recommender system and a video sentiment analysis component. The average absolute error of the system has been significantly reduced. Added system reaction to emotional feedback in the context of user interaction with virtual art compositions. Conclusion. Thus, the system can not only select the most suitable virtual art compositions, but also create adaptive and dynamic content, which will increase user satisfaction and improve the immersive aspects of the system. A promising direction of further research may be the addition of a subsystem with a generative neural network, which will create new virtual art compositions based on the conclusions of the developed recommendation system.

A Technical Review of the State-of-the-Art Methods in Aspect-Based Sentiment Analysis

With the advent and rapid advancement of text mining technology, a computer-based approach used to capture sentiment standpoints from data in textual form is increasingly becoming a promising field. Detailed information about sentiment can be provided using aspect-based sentiment analysis, which can be used in better decision-making. This study aims to study, observe, and classify previous methods used in aspect-based sentiment analysis. A systematic review is adopted as the method used to collect and review papers to achieve this research's aim. Papers focused on sentiment analysis, aspect extraction, and aspect aggregation from different academic databases such as Scopus, ScienceDirect, IEEE Explore, and Web of Science were gathered based on the inclusion and exclusion criteria of the study. The gathered papers were further reviewed to answer the stated research questions. The findings from the research show the most used methods for aspect extraction, sentiment analysis, and aspect aggregation in aspect-based sentiment analysis. This research offers a robust synthesis of evidence to guide further academic exploration in sentiment analysis.

A Comprehensive Review of Text Sentiment Analysis: A Survey of Traditional Methods and Deep Learning Approaches

Text sentiment analysis is a crucial aspect within the realm of natural language processing. This paper, incorporating the latest research advancements, systematically reviews and summarizes mainstream methods in text sentiment analysis. It covers a spectrum of traditional text analysis techniques, including rule-based and dictionary-based methods, as well as machine learning approaches. The paper further delves into an in-depth discussion of text sentiment analysis methods based on deep learning. This encompasses Recurrent Neural Networks and their enhanced structures, Convolutional Neural Networks and their extended techniques, along with applications involving pre-trained models and transfer learning.