Sentiment Lexicon Research Articles

Lexicon-Based Approach in Sentiment Analysis of Yemeni Dialect for Social Media Network

Recently, the number of Yemeni users has been expanding quickly on social media platforms. Most research in Arabic sentiment analysis has gained on Modern Standard Arabic (MSA) and some specific dialects, such as Egyptian, Levantine, and Gulf. However, there is a noticeable gap in Yemeni dialect sentiment analysis research. The reason for that is the lack of reliable Yemeni lexical and corpus and a real dataset for social media sentiment analysis. This research addresses this lack by presenting the Yemeni Dialect sentiment lexicon and corpus. This lexicon and corpus provide valuable resources for researchers and practitioners seeking to analyze sentiment in Yemeni dialect social media content, contributing to a better understanding of Yemeni public opinion, social media monitoring, marketing, cultural understanding, and assisting in efforts to respond to crises in Yemen. The Yemeni Dialect sentiment lexicon is enriched with a reasonable number of words and phrases categorized according to their positive and negative sentiment tendencies. Moreover, we constructed a corpus dataset of more than 54,000 comments built from the Facebook platform. A large dataset of unlabeled comments from the main Yemeni telecommunications companies in Yemen (Yemen Telecom, Yemen Mobile, YOU, and Sabafon), are people commenting on a public issue related to the services provided by those companies. The lexicon-based approach is used to extract the sentiment’s polarity and label each of the provided comments to formulate a corpus dataset as being either positive, negative, or neutral. The evaluation metrics of experiments are accuracy, recall, precision, f-measure, and the confusion matrix. The accuracy result of the lexicon-based labeling approach was calculated through a comparison between the achieved results and the ones achieved through manually labeled comments by three Yemeni experts. Evaluation results using a lexicon-based approach achieved an accuracy of 90.05%.

JST and rJST: joint estimation of sentiment and topics in textual data using a semi-supervised approach

ABSTRACT This paper demonstrates the performance of the Joint Sentiment Topic model (JST) and the reversed Joint Sentiment Topic model (rJST) in measuring sentiment in political speeches, comparing them against a set of popular methods for sentiment analysis: widely used off-the-shelf sentiment dictionaries; an embeddings-enhanced dictionary approach; Latent Semantic Scaling, a semi-supervised approach; and a zero-shot transformer-based approach using a large language model (GPT-4). The findings reveal JST’s superiority over all non-transformer-based approaches in predicting human-coded sentiment in multiple languages and its ability to replicate known sentiment trends in legislative speech. rJST, meanwhile, provides valuable topic-specific sentiment estimates, responsive to political dynamics and significant events. Both models are, however, outperformed by transformer-based models like GPT-4. Additionally, the paper introduces the ’sentitopics’ R-package, designed to facilitate the use of JST and rJST in computational text analysis workflows. This package is compatible with popular text analysis tools, making the models accessible for applied researchers in communication science.

Convergence or distinctiveness? Exploring how geographic location influences fan behavior in online fan communities

PurposeThis article examines how geographic location (i.e. local or nonlocal) influences fans’ communicative convergence or distinctiveness in the context of an informal online fan community (i.e. Twitter [X]). Nonlocal fans, who live geographically distant from the team they support, derive distinctiveness and belonging from supporting a team from afar. This study considers how these needs materialize in online communication.Design/methodology/approachText mining and textual analysis were employed to collect and evaluate tweets (N = 12,865) in terms of expressed emotion. Tweets were evaluated using the NRC VAD sentiment lexicon (i.e. valence, arousal, dominance) and then categorized based on geographic location and in-group/out-group status. The resulting groups were then analyzed with a two-way multivariate analysis of variance (MANOVA) to examine how geographic location relates to the expressed emotion of tweets.FindingsThe findings indicate nonlocal fans expressed statistically greater arousal compared to local online community members, reflecting their distinctiveness as nonlocal fans. Nonlocal and local fans expressed similar degrees of valence and dominance in their language, suggesting nonlocal fans converge communication behavior, seeking social approval from the salient online fan community.Originality/valueThis study suggests that geographic location influences communication behavior in online fan communities. Moreover, nonlocal fans may be valuable, and underutilized, advocates for sports teams in online community settings.

Text Analytics on YouTube Comments for Food Products

YouTube is a popular social media platform in the contemporary digital landscape. The primary focus of this study is to explore the underlying sentiment in user comments about food-related videos on YouTube, specifically within two pivotal food categories: plant-based and hedonic product. We labeled comments using sentiment lexicons such as TextBlob, VADER, and Google’s Sentiment Analysis (GSA) engine. Comment sentiment was classified using advanced Machine-Learning (ML) algorithms, namely Support Vector Machines (SVM), Multinomial Naive Bayes, Random Forest, Logistic Regression, and XGBoost. The evaluation of these models encompassed key macro average metrics, including accuracy, precision, recall, and F1 score. The results from GSA showed a high accuracy level, with SVM achieving 93% accuracy in the plant-based dataset and 96% in the hedonic dataset. In addition to sentiment analysis, we delved into user interactions within the two datasets, measuring crucial metrics, such as views, likes, comments, and engagement rate. The findings illuminate significantly higher levels of views, likes, and comments in the hedonic food dataset, but the plant-based dataset maintains a superior overall engagement rate.

A Sentiment Analysis of News Articles Published Before and During the COVID-19 Pandemic

This study investigates the impact of the COVID-19 pandemic on the connotative language used in news articles, leveraging sentiment analysis to gauge shifts in societal attitudes and potential implications for mental health. Utilizing the statistical programming language R, we extracted and analyzed texts from 645 articles published before and during the pandemic by nine authors across three major U.S. newspapers: The Wall Street Journal, New York Times, and The Washington Post. Employing the AFINN and NRC sentiment lexicons, we observed a statistically significant decrease in sentiment during the pandemic period (p < 0.0001), suggesting a pervasive shift in media discourse. This decline, consistent across newspapers and journalists, highlights the profound impact of the pandemic on societal attitudes, reflecting the pain and stress experienced by many. Such a decline in sentiment can create a negative feedback loop that exacerbates the already significant health and behavioral challenges triggered by the pandemic and its associated mitigation measures. Our findings underscore the value of sentiment analysis and text mining in assessing the effects of high-stress, long-term events on global public health while identifying a gap in the existing literature that prioritizes disease-focused research over holistic well-being. This study highlights the critical role of journalists and leaders in shaping public sentiment during crises, advocating for early recognition of concerning trends. It also offers a valuable framework for future research connecting major events with the overall media sentiment and their subsequent effects on public health.

Finding the Needle in the Haystack: Can Natural Language Processing of Students' Evaluations of Teachers Identify Teaching Concerns?

Institutions rely on student evaluations of teaching (SET) to ascertain teaching quality. Manual review of narrative comments can identify faculty with teaching concerns but can be resource and time-intensive. To determine if natural language processing (NLP) of SET comments completed by learners on clinical rotations can identify teaching quality concerns. Single institution retrospective cohort analysis of SET (n = 11,850) from clinical rotations between July 1, 2017, and June 30, 2018. The performance of three NLP dictionaries created by the research team was compared to an off-the-shelf Sentiment Dictionary. The Expert Dictionary had an accuracy of 0.90, a precision of 0.62, and a recall of 0.50. The Qualifier Dictionary had lower accuracy (0.65) and precision (0.16) but similar recall (0.67). The Text Mining Dictionary had an accuracy of 0.78 and a recall of 0.24. The Sentiment plus Qualifier Dictionary had good accuracy (0.86) and recall (0.77) with a precision of 0.37. NLP methods can identify teaching quality concerns with good accuracy and reasonable recall, but relatively low precision. An existing, free, NLP sentiment analysis dictionary can perform nearly as well as dictionaries requiring expert coding or manual creation.

Natural Language Processing (NLP) for Sentiment Analysis in Social Media

Purpose: This study sought to analyze Natural Language Processing (NLP) for sentiment analysis in social media. Methodology: The study adopted a desktop research methodology. Desk research refers to secondary data or that which can be collected without fieldwork. Desk research is basically involved in collecting data from existing resources hence it is often considered a low cost technique as compared to field research, as the main cost is involved in executive’s time, telephone charges and directories. Thus, the study relied on already published studies, reports and statistics. This secondary data was easily accessed through the online journals and library. Findings: The findings reveal that there exists a contextual and methodological gap relating to Natural Language Processing (NLP) for sentiment analysis in social media. Preliminary empirical review revealed that advanced computational techniques significantly advanced the understanding of sentiments across diverse social media platforms. Methodologies such as machine learning algorithms and deep learning models like CNNs and RNNs demonstrated robust capabilities in categorizing sentiments accurately and capturing contextual nuances such as sarcasm and slang. The research highlighted the interdisciplinary nature of NLP applications, integrating linguistics with computer science and social sciences to develop effective frameworks for analyzing large-scale social media data. These findings contributed to enhancing decision-making in marketing, politics, and public opinion research, pointing towards future directions in hybrid NLP models for improved sentiment analysis across different languages and cultural contexts. Unique Contribution to Theory, Practice and Policy: The Social Constructionism, Cognitive Linguistics and Discourse Analysis Theory may be used to anchor future studies on Natural Language Processing (NLP). The recommendations aimed to advance theoretical foundations by exploring deep learning models and nuanced sentiment lexicons. Practical applications were enhanced through the development of scalable NLP tools for real-time data processing and integration into social media platforms. Policy implications focused on establishing ethical guidelines for data privacy and bias mitigation in sentiment analysis algorithms. Cross-disciplinary collaboration fostered innovation by integrating insights from computer science, linguistics, psychology, and social sciences. Education initiatives and international collaborations were prioritized to build capacity and standardize methodologies globally, ensuring advancements in both research and practical deployment of sentiment analysis technologies.

DoSLex: automatic generation of all domain semantically rich sentiment lexicon

DoSLex: automatic generation of all domain semantically rich sentiment lexicon

Sentiment analysis technique on product reviews using Inception Recurrent Convolutional Neural Network with ResNet Transfer Learning

ABSTRACT Nowadays, online shopping has become a typical option for customers to make purchases due to the rapid technological development of the Internet. Sentiment analysis (SA) of a huge count of user reviews on e-commerce sites can successfully increase the fulfillment of user, but it is difficult so far to predict the precise sentiment polarizations of the user evaluations owing to the variations on textual arrangement, sequence length, and complex logic. In this manuscript, Inception Recurrent Convolutional Neural Network with ResNet Transfer Learning by Red Fox Optimization Algorithm using sentiment analysis on product review (IRCNN-ResNet-RFOA-SA-PR) is proposed. Here, the input data are collected through Amazon Product Reviews database and then preprocessing the input data by using whitespace tokenization, Snowball stemming, and Gensim lemmatization. The preprocessing data are supplied to the ternary pattern and discrete wavelet transform for feature extraction. Then the optimal features are selected by Fire Hawk Optimization Algorithm. After that, IRCNN-ResNet recommends the product based on SA. Finally, IRCNN-ResNet parameter is optimized by RFOA. The proposed method attains 22.37%, 31.08%, and 21.90% greater accuracy and 19.37%, 21.08%, and 25.40% greater precision when compared with existing models, such as machine learning-based SA-PR including new term weighting with feature selection mode, weighted word embedding with deep neural network-based SA-PR, and SA for e-commerce PRs in Chinese utilizing sentiment lexicon with deep learning, respectively.

Lexicon-Enhanced Long Short-Term Memory (LSTM) for Detecting of Fake News using Deep Learning

Due to its increasing popularity, low cost, and easy-to-access nature, online social media (OSM) networks have evolved as a powerful platform for people to access, consume, and share news.However, this has led to the large-scale distribution of fake news, i.e., deliberate, false, or misleading information. Spreading fake news is roughly as dangerous as spreading the virus. Fake news detection attracts many researchers' attention due to the negative impacts on the society Over the past years, many fake news detection approaches have been introduced, and most of the existing methods rely on either news content or the social context of the news dissemination process on social media platforms. In this work, we propose a lexicon-enhanced LSTM an automated model that is able to take into account both the news content and the social context for the identification of fake news. The model first uses sentiment lexicon as an extra information pre-training a word sentiment classifier and then get the sentiment embeddings of words including the words not in the lexicon. Combining the sentiment embedding and its word embedding can make word representation more accurate and to detect fake news and better predict fake user accounts and posts.We used five performance metrics to evaluate the proposed framework: accuracy, the area under the curve, precision, recall, and f1-score.The model achieves an accuracy of 99.55% compared to 93.62% against discourse structure analysis. Also, it shows an average improvement of 18.76% against other approaches, which indicates its viability against fake-classifier-based models

Research on Emotion Analysis Based on a Topic Emotion Fusion Model: The Case of the "Post-holiday Symptoms of Pharyngitis" Event on the Weibo Platform

[Objective] From early 2023 to 2024, multiple public health incidents related to "post-holiday pharyngitis" erupted in china, impacting public health and education among other areas. This study analyzes the data dissemination trends and the emotional development of these incidents. The insights are aimed at enhancing the government and disease control agencies' mechanisms for guiding and responding to such public sentiment incidents, thereby improving preparedness and response capabilities for potential future public health events. [Methods] The methodology employed involves using three instances of "post-holiday pharyngitis" events on the Weibo platform as data samples. The study applies the LDA model integrated with TD-IDF weighted Word2vec algorithm for thematic organization across different periods. The DUTIR sentiment dictionary is used to analyze public sentiment, and data visualization techniques are employed to present the thematic-emotional evolution within the incidents. [Results] The results suggest that the proposed thematic-emotional fusion model effectively extracts and analyzes the dissemination trends and emotional evolution of the event, providing actionable suggestions for government and public health agencies in managing such public sentiment incidents.

Construction of the tourist sentiment dictionary for hotels to mining tourist demands: Based on Macao’s hotel reviews from Agoda

<p><span lang="AZ-LATIN">Tourist hotels (or tourist accommodations) are located near tourist attractions, primarily serving tourists. In recent years, with the gradual improvement of people’s living standards around the globe, tourists’ demands and standards for tourist hotel construction have been rising accordingly. In the context of technologization and informatization, various hotel booking platforms (Agoda, Booking, Trip, etc.) cover a large amount of review data in evaluating systems to reflect tourists’ demands. Meanwhile, identifying demand-oriented reviews and extracting core consumer demands from them is crucial for optimizing hotel services and enhancing tourist satisfaction. Therefore, this study explores the demands of tourists in tourist hotels from the perspective of text sentiment analysis and takes Macao, a famous tourist destination, as an example, based on reviews of tourist hotels on the Agoda site platform. Specifics are as follows: (1) Based on pointwise mutual information (PMI) and information entropy (IE), it realizes the identification of sentiment words in the field of tourist hotels and constructs a sentiment dictionary to address the problem of poor relevance between word segmentation results; (2) It summarizes the five types of reviews containing tourist demands (positive, negative, suggestion, demand, and comparison) and their characteristics to solve the ambiguity of texts and further accurately reveal the main demands of tourists; (3) It classifies tourist demands and group similar tourist demands into the same categories to address the problem of multiple expressions for the same demand. The present study provides empirical experiences from Macao’s hotels and contributes to the literature on text sentiment analysis in tourist hotels. Furthermore, the study results could enhance the mining accuracy and provide a detailed summarization of consumer demands and directions for the sustainable optimization improvement of tourist services.</span></p>

Exploring the diversity of emotion in hospitality and tourism from big data: a novel sentiment dictionary

PurposeAlthough the importance and variety of emotions have been emphasized in existing literature, studies on discrete emotions remain limited. This study aims to propose a method for more precise recognition and calculation of emotions in massive amounts of online data on attraction visitor experiences and behaviour, by using discrete emotion theory.Design/methodology/approachUsing HowNet’s word similarity calculation technique, this study integrated multiple generic dictionaries, including the sentiment vocabulary ontology database of the Dalian University of Technology, the National Taiwan University Sentiment Dictionary and the Boson Dictionary. Word2vec algorithm filters emotion words unique to hospitality and tourism in 1,596,398 texts from Sogou News, Wikipedia and Ctrip reviews about attractions, and 1,765,691 reviews about attractions in China.FindingsThe discrete sentiment dictionary developed in this study outperformed the original dictionary in identifying and calculating emotions, with a total vocabulary extension of 12.07%, demonstrating its applicability to tourism.Research limitations/implicationsThe developed new dictionary can be used by researchers and managers alike to quickly and accurately evaluate products and services based on online visitor reviews.Originality/valueTo the best of the authors’ knowledge, this study is the first to construct a sentiment dictionary based on discrete emotion theory applicable to hospitality and tourism in the Chinese context. This study extended the applicability of affective psychology to hospitality and tourism using discrete emotion theory. Moreover, the study offers a methodological framework for developing a domain-specific sentiment dictionary, potentially applicable to other domains in hospitality.

Construction of the tourist sentiment dictionary for hotels to mining tourist demands: Based on Macao’s hotel reviews from Agoda

<p><span lang="AZ-LATIN">Tourist hotels (or tourist accommodations) are located near tourist attractions, primarily serving tourists. In recent years, with the gradual improvement of people’s living standards around the globe, tourists’ demands and standards for tourist hotel construction have been rising accordingly. In the context of technologization and informatization, various hotel booking platforms (Agoda, Booking, Trip, etc.) cover a large amount of review data in evaluating systems to reflect tourists’ demands. Meanwhile, identifying demand-oriented reviews and extracting core consumer demands from them is crucial for optimizing hotel services and enhancing tourist satisfaction. Therefore, this study explores the demands of tourists in tourist hotels from the perspective of text sentiment analysis and takes Macao, a famous tourist destination, as an example, based on reviews of tourist hotels on the Agoda site platform. Specifics are as follows: (1) Based on pointwise mutual information (PMI) and information entropy (IE), it realizes the identification of sentiment words in the field of tourist hotels and constructs a sentiment dictionary to address the problem of poor relevance between word segmentation results; (2) It summarizes the five types of reviews containing tourist demands (positive, negative, suggestion, demand, and comparison) and their characteristics to solve the ambiguity of texts and further accurately reveal the main demands of tourists; (3) It classifies tourist demands and group similar tourist demands into the same categories to address the problem of multiple expressions for the same demand. The present study provides empirical experiences from Macao’s hotels and contributes to the literature on text sentiment analysis in tourist hotels. Furthermore, the study results could enhance the mining accuracy and provide a detailed summarization of consumer demands and directions for the sustainable optimization improvement of tourist services.</span></p>

Optimizing Sentiment Analysis Models for Customer Support: Methodology and Case Study in the Portuguese Retail Sector

Sentiment analysis is a cornerstone of natural language processing. However, it presents formidable challenges due to the intricacies of lexical diversity, complex linguistic structures, and the subtleties of context dependence. This study introduces a bespoke and integrated approach to analyzing customer sentiment, with a particular emphasis on a case study in the Portuguese retail market. Capitalizing on the strengths of SentiLex-PT, a sentiment lexicon curated for the Portuguese language, and an array of sophisticated machine learning algorithms, this research constructs advanced models that encapsulate both lexical features and the subtleties of linguistic composition. A meticulous comparative analysis singles out multinomial logistic regression as the pre-eminent model for its applicability and accuracy within our case study. The findings of this analysis highlight the pivotal role that sentiment data play in strategic decision-making processes such as reputation management, strategic planning, and forecasting market trends within the retail sector. To the extent of our knowledge, this work is pioneering in its provision of a holistic sentiment analysis framework tailored to the Portuguese retail context, marking an advancement for both the academic field and industry application.

The difference in positivity of the Russian and English lexicon: The big data approach

Psychological cross-cultural studies have long noted differences in the degree of cognition positivity, or optimism, in various cultures. Herewith, the question whether the difference shows up at the level of the language lexicon remains unexplored. Linguistic positivity bias has been confirmed for a number of languages. The point of it is that most words have a positive connotation in the language. This begs the question: is linguistic positivity bias the same for different languages or not? In a sense, the issue is similar to the hypothesis of linguistic relativity suggesting the language impact on the human cognitive system. The problem has been researched only in one work (Dodds et al. 2015), where data on the positivity bias values are given for different languages and the comparison for each pair of languages is based on merely one pair of dictionaries. In the present study, we radically increase the computational baseline by comparing four English and five Russian dictionaries. We carry out the comparative study both at the level of vocabularies and at the level of texts of different genres. A new, previously untapped idea is to compare positivity ratings of translated texts. Also, English and Russian sentiment dictionaries are compared based on the scores of translation-stable words. The results suggest that the Russian language is somewhat slightly more positive than English at the level of vocabulary.

Exploring the Use of Natural Language Processing to Understand Emotions of Trainees and Faculty Regarding Entrustable Professional Activity Assessments.

Background In medical education, artificial intelligence techniques such as natural language processing (NLP) are starting to be used to capture and analyze emotions through written text. Objective To explore the application of NLP techniques to understand resident and faculty emotions related to entrustable professional activity (EPA) assessments. Methods Open-ended text data from a survey on emotions toward EPA assessments were analyzed. Respondents were residents and faculty from pediatrics (Peds), general surgery (GS), and emergency medicine (EM), recruited for a larger emotions study in 2023. Participants wrote about their emotions related to receiving/completing EPA assessments. We analyzed the frequency of words rated as positive via a validated sentiment lexicon used in NLP studies. Specifically, we were interested if the count of positive words varied as a function of group membership (faculty, resident), specialty (Peds, GS, EM), gender (man, woman, nonbinary), or visible minority status (yes, no, omit). Results A total of 66 text responses (30 faculty, 36 residents) contained text data useful for sentiment analysis. We analyzed the difference in the count of words categorized as positive across group, specialty, gender, and being a visible minority. Specialty was the only category revealing significant differences via a bootstrapped Poisson regression model with GS responses containing fewer positive words than EM responses. Conclusions By analyzing text data to understand emotions of residents and faculty through an NLP approach, we identified differences in EPA assessment-related emotions of residents versus faculty, and differences across specialties.

The social media sentiment analysis framework: deep learning for sentiment analysis on social media

Researching public opinion can help us learn important facts. People may quickly and easily express their thoughts and feelings on any subject using social media, which creates a deluge of unorganized data. Sentiment analysis on social media platforms like Twitter and Facebook has developed into a potent tool for gathering insights into users' perspectives. However, difficulties in interpreting natural language limit the effectiveness and precision of sentiment analysis. This research focuses on developing a social media sentiment analysis (SMSA) framework, incorporating a custom-built emotion thesaurus to enhance the precision of sentiment analysis. It delves into the efficacy of various deep learning algorithms, under different parameter calibrations, for sentiment extraction from social media. The study distinguishes itself by its unique approach towards sentiment dictionary creation and its application to deep learning models. It contributes new insights into sentiment analysis, particularly in social media contexts, showcasing notable advancements over previous methodologies. The results demonstrate improved accuracy and deeper understanding of social media sentiment, opening avenues for future research and applications in diverse fields.

Sentibank: A Unified Resource of Sentiment Lexicons and Dictionaries

Sentibank: A Unified Resource of Sentiment Lexicons and Dictionaries

Automatic construction of direction-aware sentiment lexicon using direction-dependent words

AbstractExplainability, which is the degree to which an interested stakeholder can understand the key factors that led to a data-driven model’s decision, has been considered an essential consideration in the financial domain. Accordingly, lexicons that can achieve reasonable performance and provide clear explanations to users have been among the most popular resources in sentiment-based financial forecasting. Since deep learning-based techniques have limitations in that the basis for interpreting the results is unclear, lexicons have consistently attracted the community’s attention as a crucial tool in studies that demand explanations for the sentiment estimation process. One of the challenges in the construction of a financial sentiment lexicon is the domain-specific feature that the sentiment orientation of a word can change depending on the application of directional expressions. For instance, the word “cost” typically conveys a negative sentiment; however, when the word is juxtaposed with “decrease” to form the phrase “cost decrease,” the associated sentiment is positive. Several studies have manually built lexicons containing directional expressions. However, they have been hindered because manual inspection inevitably requires intensive human labor and time. In this study, we propose to automatically construct the “sentiment lexicon composed of direction-dependent words,” which expresses each term as a pair consisting of a directional word and a direction-dependent word. Experimental results show that the proposed sentiment lexicon yields enhanced classification performance, proving the effectiveness of our method for the automated construction of a direction-aware sentiment lexicon.