The Role of User-Generated Content Data for Collaborative Learning: Identifying Tourism Hot Topics During the Pandemic

Lockdowns and physical distancing measures due to COVID-19 have entailed an unprecedent disruption on the tourism industry, which has resulted in the cancelation of many in-person trips in the wake of the pandemic. In view of COVID-19 spread, many worldwide tourism companies have been deciding what measures to adopt so as to limit face to face contacts. Besides, tourists are seeking services that fulfil sanitary protocols as well as experiences that cheer them up. Social media has become not just an entertainment tool but rather a socializing channel that is employed on a daily basis by individuals, companies, organizations, governments due to the vast benefits they get from it. Twitter has been identified as the most popular microblog platform, and a reliable source for examining and studying the industry's behavior. It is unknown which tourism topics have been predominantly discussed in social media. The discussion identifies the major themes using user-generated content published on Twitter. A dataset of tweets was employed to classify the tourism categories most discussed from the 30 November 2020 to 25 January 2021, using a Latent Dirichlet Allocation model. We applied sentiment analysis using machine learning in Python to distinguish between the positive, negative and neutral feelings expressed in the tweets. The goal is to explore of these sentiments coincide with sustainable development goals, boost collaborative learning and expand tourism user-generated content during a health crisis. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022.

The Investigation into Deep Learning Classifiers Towards Imbalanced Text Data

Class imbalance is an important classification problem where failure to identify events can be hazardous due to failure of solution preparation or opportune handling. Minorities are mostly more consequential in such cases. It is necessary to know a reliable classifier for imbalanced classes. This study examines several conventional machine learning and deep learning methods to compare the performance of each method on dataset with imbalanced classes. We use COVID-19 online news titles to simulate different class imbalance ratios. The results of our study demonstrate the superiority of the CNN with embedding layer method on a news titles dataset of 16,844 data points towards imbalance ratios of 37%, 30%, 20%, 10%, and 1%. However, CNN with embedding layer showed a noticeable performance degradation at an imbalance ratio of 1%. © 2022 IEEE.

Contextualizing Emerging Trends in Financial News Articles

Identifying and exploring emerging trends in news is becoming more essential than ever with many changes occurring around the world due to the global health crises. However, most of the recent research has focused mainly on detecting trends in social media, thus, benefiting from social features (e.g. likes and retweets on Twitter) which helped the task as they can be used to measure the engagement and diffusion rate of content. Yet, formal text data, unlike short social media posts, comes with a longer, less restricted writing format, and thus, more challenging. In this paper, we focus our study on emerging trends detection in financial news articles about Microsoft, collected before and during the start of the COVID-19 pandemic (July 2019 to July 2020). We make the dataset accessible and we also propose a strong baseline (Contextual Leap2Trend) for exploring the dynamics of similarities between pairs of keywords based on topic modeling and term frequency. Finally, we evaluate against a gold standard (Google Trends) and present noteworthy real-world scenarios regarding the influence of the pandemic on Microsoft. ©2022 Association for Computational Linguistics.

Understanding Desire to Touch Using Large-scale Twitter Data

I and research colleagues investigated people's desire to touch by collecting and analyzing a large amount of text data that contain phrases such as "want to touch” on Twitter. We revealed the relationship between the body part that people want to touch and the touch gesture. We also revealed the effects of the COVID-19 pandemic on the desire to touch. Specifically, we observed "skin hunger,” i.e., the strong desire for physical communication, and variation of touch avoidance toward objects such as doorknobs. Our results will be beneficial for understanding human behavior as well as for the further development of haptic technology. © 2023 Nippon Telegraph and Telephone Corp.. All rights reserved.

Social Media Analysis using Sentiment Analysis on COVID-19 from Twitter

The development of the internet is getting faster, participating in encouraging the emergence of new and innovative information. In filtering the various information that appears, we need a recommended system to perform well for users in today's internet era. A well-performing recommendation system in question is a reliable recommendation algorithm. This algorithm is fundamental to analyzing various information, such as responses on social media based on user behavior data related to the topic of COVID. This data is crawled from tweets on social media Twitter. The data analysis algorithm obtained uses Python, which is then visualized in the form of a diagram. The processed data is user comments on Twitter, and the text data is analyzed using Python, using more than 60000 data sets taken to form visualizations and conclusions. From sentiment analysis, polarity and subjectivity data are obtained to be analyzed, which are negative, neutral, or positive. The result is show positive tweets with 29.2%, negative tweets is 13%, and 57.8% neutral tweets. Lastly, sentiment analysis can help people effectively infer vast and complex data from social media like Twitter. © 2022 IEEE.

Sentiment and Mobility Analysis on COVID-19 Restrictions with Autoencoder

In 2020, the world was attacked by a virus known as the COVID-19 virus. Restrictions on people's activities were conducted in various countries to prevent the spread of the virus. However, since people were vaccinated, restriction levels have been reduced or eliminated, although the new cases of COVID-19 worldwide have not ended. People's responses to restriction policies vary, including sentiment and human mobility. The possibility of sentiment is either support or resistance, while mobility is staying at home or not. This study analyzes the proportion between the two responses through two types of data: Text for sentiment and time series for mobility. Sentiment text data is taken from Twitter and mobility time series data is taken from Google Mobility for February 2020 to April 2022. Twitter and Google Mobility data are collected from several countries using English and implementing restrictions: Australia, Canada, Singapore, the United Kingdom (UK), and the United States (US). The unsupervised Autoencoder model is leveraged to find clusters. Two Autoencoder architectures are proposed for each data type. Before being used in Multilayer Autoencoder, text data is converted to vector data by Word2Vec. On the other hand, LSTM-Autoencoder is used for time series data. Finally, hypothesis tests are performed to determine the mean between the clusters formed is the same or different, out of five countries, only Canada has a null hypothesis is accepted, that means people in Canada tend to be neutral in response to COVID-19 while mobilities are dynamics, it reveals that people in Canada obey the government's decision on restrictions during the rise of COVID-19 cases. © 2022 ACM.

Deep Learning in the Detection of Disinformation about COVID-19 in Online Space.

This article focuses on the problem of detecting disinformation about COVID-19 in online discussions. As the Internet expands, so does the amount of content on it. In addition to content based on facts, a large amount of content is being manipulated, which negatively affects the whole society. This effect is currently compounded by the ongoing COVID-19 pandemic, which caused people to spend even more time online and to get more invested in this fake content. This work brings a brief overview of how toxic information looks like, how it is spread, and how to potentially prevent its dissemination by early recognition of disinformation using deep learning. We investigated the overall suitability of deep learning in solving problem of detection of disinformation in conversational content. We also provided a comparison of architecture based on convolutional and recurrent principles. We have trained three detection models based on three architectures using CNN (convolutional neural networks), LSTM (long short-term memory), and their combination. We have achieved the best results using LSTM (F1 = 0.8741, Accuracy = 0.8628). But the results of all three architectures were comparable, for example the CNN+LSTM architecture achieved F1 = 0.8672 and Accuracy = 0.852. The paper offers finding that introducing a convolutional component does not bring significant improvement. In comparison with our previous works, we noted that from all forms of antisocial posts, disinformation is the most difficult to recognize, since disinformation has no unique language, such as hate speech, toxic posts etc.

Visual Analytics-Based Method for Sentiment Analysis of COVID-19 Ukrainian Tweets

The coronavirus pandemic is one of the leading communication topics for users on social networks. It causes different emotions in people: fear, sadness, anger, joy, and elation. Detecting sentiment about the pandemic is an acute challenge because it helps track people’s attitudes about the pandemic itself and the messages and decisions of local authorities aimed at combating the coronavirus. To address the issue, namely natural language processing, messages are processed using the TextRank vectorization method and the SVM-based two-level classification model. The first stage is the detection of tweets that are directly related to the coronavirus. The second stage means detecting the sentiment of the dataset obtained in the first stage. The classifier’s effectiveness was tested using the following metrics: precision, recall, F1-norm, and confusion matrix, and averaged about 90%. Thus, the automated detection of the sentiment of Twitter messages about the coronavirus pandemic was obtained. The approach described in the paper will allow assessing public opinion on pandemic control measures applied by the country’s governments. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Machine Learning and Lexicon Approach to Texts Processing in the Detection of Degrees of Toxicity in Online Discussions.

This article focuses on the problem of detecting toxicity in online discussions. Toxicity is currently a serious problem when people are largely influenced by opinions on social networks. We offer a solution based on classification models using machine learning methods to classify short texts on social networks into multiple degrees of toxicity. The classification models used both classic methods of machine learning, such as naïve Bayes and SVM (support vector machine) as well ensemble methods, such as bagging and RF (random forest). The models were created using text data, which we extracted from social networks in the Slovak language. The labelling of our dataset of short texts into multiple classes-the degrees of toxicity-was provided automatically by our method based on the lexicon approach to texts processing. This lexicon method required creating a dictionary of toxic words in the Slovak language, which is another contribution of the work. Finally, an application was created based on the learned machine learning models, which can be used to detect the degree of toxicity of new social network comments as well as for experimentation with various machine learning methods. We achieved the best results using an SVM-average value of accuracy = 0.89 and F1 = 0.79. This model also outperformed the ensemble learning by the RF and Bagging methods; however, the ensemble learning methods achieved better results than the naïve Bayes method.

New Frontiers of Scientific Text Mining: Tasks, Data, and Tools

Exploring the vast amount of rapidly growing scientific text data is highly beneficial for real-world scientific discovery. However, scientific text mining is particularly challenging due to the lack of specialized domain knowledge in natural language context, complex sentence structures in scientific writing, and multi-modal representations of scientific knowledge. This tutorial presents a comprehensive overview of recent research and development on scientific text mining, focusing on the biomedical and chemistry domains. First, we introduce the motivation and unique challenges of scientific text mining. Then we discuss a set of methods that perform effective scientific information extraction, such as named entity recognition, relation extraction, and event extraction. We also introduce real-world applications such as textual evidence retrieval, scientific topic contrasting for drug discovery, and molecule representation learning for reaction prediction. Finally, we conclude our tutorial by demonstrating, on real-world datasets (COVID-19 and organic chemistry literature), how the information can be extracted and retrieved, and how they can assist further scientific discovery. We also discuss the emerging research problems and future directions for scientific text mining. © 2022 Owner/Author.

Modeling Student Discourse in Online Discussion Forums Using Semantic Similarity Based Topic Chains

Students’ conversations in academic settings evolve over time and can be affected by events such as the COVID-19 pandemic. In this paper, we employ a Contextualized Topic Modeling technique to detect coherent topics from students’ posts in online discussion forums. We construct topic chains by connecting semantically similar topics across months using Word Mover’s Distance. Consistent academic discourse and contemporary events such as the COVID-19 outbreak and the Black Lives Matter movement were found among prominent topics. In later months, new themes around students’ lived experiences emerged and evolved into discussions reflecting the shift in educational experiences. Results revealed a significant increase in more general topics after the onset of pandemic. Our proposed framework can also be applied to other contexts investigating temporal topic trends in large-scale text data. © 2022, Springer Nature Switzerland AG.

A Digital Forensic Readiness Cybercrime Semantic Trigger Process

The recent wave of the global Covid-19 pandemic has led to a surge in text-based non-technical cybercrime attacks within the cyber ecosystem. Information about such cyber-attacks is often in unstructured text data and metadata, a rich source of evidence in a digital forensic investigation. However, such information is usually unavailable during a digital forensic investigation when dealing with the public cloud post-incident. Furthermore, digital investigators are challenged with extracting meaningful semantic content from the raw syntactic and unstructured data. It is partly due to the lack of a structured process for forensic data pre-processing when or if such information is identified. Thus, this study seeks to address the lack of a procedure or technique to extract semantic meaning from text data of a cybercrime attack that could be used as a digital forensic readiness semantics trigger in a cybercrime detection process. For the methodology to address the proposed approach, data science modelling and unsupervised machine learning are used to design a strategy. This method process extracts tokens of cybercrime text data, which are further used to develop an intelligent DFR semantic tool extractor based on natural language patterns from cybercrime text data. The proposed DFR cybercrime semantic trigger process when implemented could be used to create a digital forensic cybercrime language API for all digital forensic investigation systems or tools. © 2022, ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.

A topic modeling method for analysis of short-text data in social media networks

Currently, many short texts are published online, especially on social media platforms. High impact events, for example, are highly commented on by users. Understanding the subjects and patterns hidden in online discussions is a very important task for contexts such as elections, natural disasters or major sporting events. However, many works of this nature use techniques that, despite showing satisfactory results, are not the most suitable when it comes to the short texts on social media and may suffer a loss in their results. Therefore, this paper presents a text mining method for messages published on social media, with a data pre-processing step and topic modeling for short texts. For this paper, we created a data set from real world tweets related to COVID-19 that is openly available1 for research purposes. © 2022, EasyChair. All rights reserved.

Analysis public opinion of the COVID-19 pandemic based on online post text data

The aim of this study is to explore the changes in the attention of the public and the trend of emotional changes in the face of the COVID-19 pandemic, based on web crawlers and big data statistical analysis technology. Public opinion information related to COVID-19 was especially obtained from Chinese online forums. The data was cleaned and filtered through data preprocessing operations, and statistical analysis was performed on multiple data characteristics to grasp the public opinion trends on the pandemic. The results of this study are of great significance for maintaining the social stability of China, the stable development of the economy, and the implementation of regular epidemic prevention and control measures and fully advancing work resumption. © 2022 IEEE.

A Machine Learning Approach to Track COVID-19 Pandemic using Sentiment Analysis

Coronavirus disease or COVID-19 is one of the most frightening and infectious diseases of the twenty-first century. Since the outbreak of COVID-19 in Wuhan, China, numerous researches are conducted in this sector. At the preliminary stage, there was not sufficient numeric data for research but when we consider the text data such as trending topics of Social Media or patients sharing experiences about their symptoms, we get enough data to ace the navigation of the Coronavirus (SARS-CoV-2). Keeping aside the health complications related to COVID-19, there also has been huge public panic following the pandemic. Sentiment analysis helps to learn the emotions of a vast number of people about any particular topic. In this paper, we have used sentiment analysis methods to observe the public reaction to the COVID-19 pandemic and people's experience of the ongoing vaccination process. Machine Learning-based (ML-based) classification algorithms are implemented for text classification. Finally, the accuracy of the classification models is also calculated for further prediction. © 2021 IEEE.

Sensing the Impact of COVID-19 Restrictions from Online Reviews: The Cases of London and Paris Unveiled Through Text Mining

This study aims to understand how the COVID-19 pandemic affected the hotel sector and to identify the current traveler demands. The traveler’s reviews were analyzed based on sentiment analysis and a guest satisfaction model was also proposed, demonstrating a data mining approach within tourism and hospitality research. Given its popularity, TripAdvisor was the chosen platform for collection of hotel reviews in London and Paris. Text data were extracted from reviews made in two time periods, before and during the COVID-19 pandemic. The sentiment and specific aspects highlighted by travelers were compared between each period. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Dynamic Topic Tracking and Visualization using COVID-19 related Tweets in Multiple Languages

With the increasing proliferation of mobile phone, internet and communication technologies, social network sites (SNS) are gaining importance worldwide. People express and exchange their opinion on various social network sites like twitter, facebook, blogs over different local and global issues. Efficient analysis of such vast text data from SNS provides a good way of understanding insights of public opinion, government policy and social condition of different countries. Topic modeling is a popular tool for extracting information from text data. Dynamic topic tracking and its visualization provides a means for capturing the change of topics over time which is important for visualization of changing needs of the society and keeping updated with the current situation. In this work, COVID-19 related twitter data in two different languages are collected and analyzed by dynamic topic model to track the spread of the evolved topics during the pandemic in two different countries in order to visualize the differences and commonness of the effect of pandemic. Here we mainly focused on the tweet data related to Japan and India in Japanese and English respectively. It is found that the country specific characteristics are prominent in some topics while some topics express the general concerns during the pandemic. This study seems to be effective to provide a technique for capturing the opinion and needs of people during a pandemic by analysis of tweet data. © 2021 IEEE.

Predicting Depression from Social Networking Data using Machine Learning Techniques

In this era of COVID-19 pandemic, as more people self-isolate themselves, psychological health issues like depression, anxiety, and stress is an increasing concern all over the world. The purpose of this study is to investigate the data from social forums, where we found communities of depressed people sharing their thoughts and emotions in the forums, these forums also receive advices and support. In this paper, we will analyse the "depressed"text;by manipulating the data, extracting features, categorising, and try to understand what are the attributes of "depressed"text, and how we can "predict"whether a text should be marked as depressed or not. Using text analysis and text data mining techniques, the text obtained from the social forums was analysed and three different machine learning algorithms were used to predict depression. After cross validation overall accuracy of 99.69% was obtained as the best score using the proposed system. This study definitively answers the question regarding using human basic language and communication of personal experiences, for the prediction of depression and can be reached easily. Furthermore, not only actions, habits and behaviour of a person, text too can be used for accurate diagnosis of depression. © 2021 IEEE.

Research evidence for reshaping global energy strategy based on trend-based approach of big data analytics in the corona era

This study is designed to discover current energy-related research trends as evidence inherent in big data and obtain future agendas and new insights to reshape global energy strategy considering environmental change globally in the era of COVID-19. To this end, vast amounts of unstructured text data on energy demand from 4 top journals in the energy field over the past year (2020–2021) were collected. This study used a semantic network analysis in big data analytics. As a result, this study obviously shows that research evidence on traditional energy sources such as fossil fuels and gas is still essential in the current situation where climate change and global warming are intensifying worldwide. Simultaneously, research on renewable energy is positioned as a critical agenda by providing sufficient evidence to draw practical implications concerning overall global energy companies' strategy and each government's new energy policy in line with the rapid changes in the global environment. Consequently, this study proposes that a decision-maker or leader can exert the remarkable power of reshaping global energy strategy for future sustainability, taking into account the recognizable trends inherent in big data. © 2022 The Author

Extractive Text Summarization for COVID-19 Medical Records

The method of reducing information from an original text document while maintaining the vital information is known as text summarizing. The amount of text data available has increased dramatically in recent years from a variety of sources. A large volume of text is an excellent source of information and knowledge of the source is essential for efficiently summarizing information that must be useful. Summarization facilitates the acquisition of vital and required information in a short period of time. Text summarization is required in a variety of domains, including news article summaries, email summaries and information summaries in the medical profession to track a patient's medical history for future treatment and so on. In summarization, there are two methods: extractive summarization and ive summarization. In this work, extractive summarization is used on the COVID-19 dataset. Different models and their results have been discussed. © 2021 IEEE.