Twitter Dataset Research Articles

Countering Social Media Cybercrime Using Deep Learning: Instagram Fake Accounts Detection

An online social media platform such as Instagram has become a popular communication channel that millions of people are using today. However, this media also becomes an avenue where fake accounts are used to inflate the number of followers on a targeted account. Fake accounts tend to alter the concepts of popularity and influence on the Instagram media platform and significantly impact the economy, politics, and society, which is considered cybercrime. This paper proposes a framework to classify fake and real accounts on Instagram based on a deep learning approach called the Long Short-Term Memory (LSTM) network. Experiments and comparisons with existing machine and deep learning frameworks demonstrate considerable improvement in the proposed framework. It achieved a detection accuracy of 97.42% and 94.21% on two publicly available Instagram datasets, with F-measure scores of 92.17% and 89.55%, respectively. Further experiments on the Twitter dataset reveal the effectiveness of the proposed framework by achieving an impressive accuracy rate of 99.42%.

Entity Linking for real-time geolocation of natural disasters from social network posts.

When a fast kinetic natural disaster occurs, it is crucial that crisis managers quickly understand the extent of the situation, especially through the development of "big picture" maps. For many years, great efforts have been made to use social networks to help build this situational awareness. While there are many models for automatically extracting information from posts, the difficulty remains in detecting and geolocating this information on the fly so that it can be placed on maps. Whilst most of the work carried out to date on this subject has been based on data in English, we tackle the problem of detecting and geolocating natural disasters from French messages posted on the Twitter platform (now renamed "X"). To this end, we first build an appropriate dataset comprised of documents from the French Wikipedia corpus, the dataset from the CAp 2017 challenge, and a homemade annotated Twitter dataset extracted during French natural disasters. We then developed an Entity-Linking pipeline in adequacy with our end-application use case: real-time prediction and peak resiliency. We show that despite these two additional constraints, our system's performances are on par with state-of-the-art systems. Moreover, the entities geolocated by our model show a strong coherence with the spatiotemporal signature of the natural disasters considered, which suggests that it could usefully contribute to automatic social network analysis for crisis managers.

Rumor Detection With Hierarchical Representation on Bipartite Ad Hoc Event Trees.

The rapid growth of social media has caused tremendous effects on information propagation, raising extreme challenges in detecting rumors. Existing rumor detection methods typically exploit the reposting propagation of a rumor candidate for detection by regarding all reposts to a rumor candidate as a temporal sequence and learning semantics representations of the repost sequence. However, extracting informative support from the topological structure of propagation and the influence of reposting authors for debunking rumors is crucial, which generally has not been well addressed by existing methods. In this article, we organize a claim post in circulation as an ad hoc event tree, extract event elements, and convert it into bipartite ad hoc event trees in terms of both posts and authors, i.e., author tree and post tree. Accordingly, we propose a novel rumor detection model with hierarchical representation on the bipartite ad hoc event trees called BAET. Specifically, we introduce word embedding and feature encoder for the author and post tree, respectively, and design a root-aware attention module to perform node representation. Then we adopt the tree-like RNN model to capture the structural correlations and propose a tree-aware attention module to learn tree representation for the author tree and post tree, respectively. Extensive experimental results on two public Twitter datasets demonstrate the effectiveness of BAET in exploring and exploiting the rumor propagation structure and the superior detection performance of BAET over state-of-the-art baseline methods.

Use of Natural Language Processing for the Detection of Hate Speech on Social Media

Our society’s communication patterns have fundamentally changed as a consequence of the emergence of social media platforms. One effect of these changes is a rise in unpleasant behaviours like making rude and derogatory comments online. Speaking harshly or disrespectfully to someone in person may be difficult. However, online abuse and posting of improper material are considered to be acceptable. Hate speech has the potential to hurt a person or a group of people. Inappropriate material must be identified, in order to be filtered or banned from the web. CNN is a type of deep machine-learning model that has been suggested for such identification, because it performs better than conventional techniques in resolving text categorization problems. Our goal investigates how hate speech may be detected using NLP. In addition, a recent technique has been used in this field to a dataset. This classifier is assigned in each tweet to one of the three Twitter dataset categories of hatred, foul language, or neither. This model’s performance has been assessed with accuracy. The Naïve Bayes, the Decision Tree, KNN, Linear Regression, and the Random Forest are five algorithms that have been used. Of these, Linear Regression provided the greatest accuracy of 94%. It should be noted that when looking at each class separately, many hateful tweets have been mislabelled. It is advisable to look at the outcomes and faults in much detail, in order to comprehend the misclassification. Our analysis shows a better outcome in detecting hateful speech in social media.

Aspect-based sentiment analysis via bidirectional variant spiking neural P systems

In recent years, aspect-based sentiment classification has predominantly relied on artificial neural networks, achieving notable success. However, effectively capturing the contextual semantic relationships between aspects and content has remained challenging. To address this issue, we propose an innovative approach for aspect-based sentiment analysis. Our method uses a Bidirectional Variant Spiking Neural P System (Bi-VSNP) as a context encoding module, integrated with a two-channel mechanism to extract context and aspect features. Furthermore, Our dual attention mechanism leverages the relationship between aspects and context, allowing the model to focus on key sentiment information. We rigorously evaluated our method on publicly available datasets, demonstrating its superior performance over most models based on artificial neural networks as context extraction. Compared to other models utilizing SNP methods, our approach achieves performance improvements of 1.72%, 1.27%, and 1.05% on the Laptop, Restaurant, and Twitter datasets, respectively. These results underscore the efficacy of our approach in aspect-based sentiment classification.

Aspect-based sentiment analysis: A dual-task learning architecture using imbalanced maximized-area under the curve proximate support vector machine and reinforcement learning

Aspect-based sentiment analysis (ABSA) deals with understanding sentiments about specific attributes. Aspect term extraction (ATE) plays an essential role in aspect sentiment classification (ASC) by specifying the main aspects for sentiment accuracy. This paper introduces a dual-task framework for ATE and ASC models in which, the ATE model uses the imbalanced maximized-area under the curve proximate support vector machine (ImAUC-PSVM) method to mitigate the imbalance problem and the output of ATE serves as the reward in the training process of ASC supported by RL (reinforcement learning). The ImAUC-PSVM uses the conventional PSVM (proximate support vector machine) strength for sentiment tasks and the AUC (area under the curve) optimization for imbalance addressing. The ASC model employs multiple kernels for word embedding, with a transductive long short-term memory (TLSTM) for sentence representation that considers samples near the test point for better model refinement. The policy of the RL-based training strategy of the ASC model is represented by a multilayer perceptron (MLP) that categorizes sentences into positive, negative and neutral sentiments. In this RL context, aspect terms serve as rewards in the training process. The scope loss function (SLF) is also integrated within RL to rectify dataset imbalances. To optimize ATE and ASC models, an artificial bee colony (ABC) technique is incorporated for hyperparameter optimization. Experiments and evaluations conducted on English (Restaurant and Laptop) and Hindi datasets show the superiority of the proposed model over existing models of the literature. Moreover, transfer learning (TL) is applied to a Twitter dataset to assess its effect on ATE. The F-measures of the ATE model are obtained as 90.12 and 84.10 and the accuracies of the ASC model are calculated as 84.10 and 82.02 on Restaurant and Laptop datasets, respectively.

Stacked Classification Approach using Optimized Hybrid Deep Learning Model for Early Prediction of Behaviour Changes on Social Media

Detecting signs of suicidal thoughts on social media is paramount for preventing suicides, given the platforms' role as primary outlets for emotional expression. Traditional embedding techniques focus solely on semantic analysis and lack the sentiment analysis essential for capturing emotions. This limitation poses challenges in developing high-accuracy models. Additionally, previous studies often rely on a single dataset, further constraining their effectiveness. To overcome these challenges, this study proposes an innovative approach that integrates embedding techniques such as BERT, which offers semantic and syntactic analysis of the posts, with sentiment analysis provided by VADER scores extracted from the VADER sentiment analysis tool. The identified features are then input into the proposed optimized hybrid deep learning model, specifically the Bi-GRU and Attention incorporated with Stacked or stacking Classifier (Decision Tree, Random Forest, Gradient Boost, as the base classifier and XGBoost as meta classifier), which undergoes optimization using the grid search technique to enhance detection capabilities. In evaluations, the model achieved an impressive accuracy and F1-score of 98% on the Reddit dataset and 97% on the Twitter dataset. The research evaluates the efficacy of several machine learning models, encompassing Decision Trees, Random Forests, Gradient Boosting, and XGBoost. Moreover, it examines sophisticated models like LSTM with Attention, Bi-LSTM with Attention, and Bi-GRU with Attention, augmented with word embeddings such as BERT, MUSE, and fastText, alongside the fusion of sentiment VADER score. These results emphasize the promise of a holistic strategy that combines advanced feature embedding techniques with semantic features, showcasing a notably efficient detection of suicidal ideation on social media.

Social Network Public Opinion Analysis Using BERT-BMA in Big Data Environment

The existing social network public opinion analysis methods have problems such as poor semantic expression quality and weak detection ability in short texts. Therefore, a social network public opinion analysis method based on BERT-BMA is proposed. To normalize the comment text, the rumor text is initially transferred to a word vector matrix using the BERT (Bidirectional Encoder Representations from Transformer) model. The BiLSTM-based network architecture is subsequently employed to acquire the trace features of data transmission. Ultimately, this study employs the multi-head attention mechanism to extract feature information that is more significant in the analysis of online public opinion by mining the dependency relationships between users, resulting in increasing ability to detect public opinion emergencies. The experimental outcomes indicate that the results on the Twitter data set and Weibo dataset are superior to other comparative models.

MOHCOA: Multi-objective hermit crab optimization algorithm for feature selection in sentiment analysis of Covid-19 Twitter datasets

MOHCOA: Multi-objective hermit crab optimization algorithm for feature selection in sentiment analysis of Covid-19 Twitter datasets

Rumour detection on benchmark twitter datasets using graph neural networks with data augmentation

Rumour detection on benchmark twitter datasets using graph neural networks with data augmentation

An aspect sentiment analysis model with Aspect Gated Convolution and Dual-Feature Filtering layers

Aspect level sentiment analysis is a basic task to determine the sentiment bias based on the contextual information near the aspect words. Some sentences contain many confusing feature words due to incomplete structure or high complexity in sentiment prediction, which can easily lead to the problem that the model pays too much attention to unimportant features. Furthermore, the role of aspect-related sentiment features is not significant in the feature extraction process. The capsule network, due to its complex network structure, leads to capsule detachment when too much information is found in the dataset. Too much computational resources are easily consumed during dynamic routing, which in turn affects the model's judgement of the polarity of aspect-related sentiment. To address these issues, we propose a capsule network (AGCDFF-Caps) with aspect-gated convolution and Dual-Feature Filtering layers for aspect-level sentiment analysis. A pre-trained BERT model is used to generate a serialised representation of the text, and the semantic representation in the contextual text is enhanced by the self-attention mechanism and bi-GRU. An aspect-gating based convolutional neural network is constructed to selectively extract contextual sentiment information about aspects and discard irrelevant information. A capsule network is then used to learn the multispatial semantic features of the aspect and the text. In particular, we incorporate a Dual-Feature Filtering network structure into the capsule network structure to strengthen the interaction between the particular aspect and the context from global and local perspectives, filtering the redundant information in local semantics and global semantics. The opinion feature representations that can more accurately express the emotional tendency of the aspect are obtained. Experimental results on SemEval2014 and Twitter datasets show that the proposed hybrid network structure has superior classification performance compared to 12 advanced baseline methods.

Keep your friends close, and your enemies closer: structural properties of negative relationships on Twitter

The Ego Network Model (ENM) is a model for the structural organisation of relationships, rooted in evolutionary anthropology, that is found ubiquitously in social contexts. It takes the perspective of a single user (Ego) and organises their contacts (Alters) into a series of (typically 5) concentric circles of decreasing intimacy and increasing size. Alters are sorted based on their tie strength to the Ego, however, this is difficult to measure directly. Traditionally, the interaction frequency has been used as a proxy but this misses the qualitative aspects of connections, such as signs (i.e. polarity), which have been shown to provide extremely useful information. However, the sign of an online social relationship is usually an implicit piece of information, which needs to be estimated by interaction data from Online Social Networks (OSNs), making sign prediction in OSNs a research challenge in and of itself. This work aims to bring the ENM into the signed networks domain by investigating the interplay of signed connections with the ENM. This paper delivers 2 main contributions. Firstly, a new and data-efficient method of signing relationships between individuals using sentiment analysis and, secondly, we provide an in-depth look at the properties of Signed Ego Networks (SENs), using 9 Twitter datasets of various categories of users. We find that negative connections are generally over-represented in the active part of the Ego Networks, suggesting that Twitter greatly over-emphasises negative relationships with respect to “offline” social networks. Further, users who use social networks for professional reasons have an even greater share of negative connections. Despite this, we also found weak signs that less negative users tend to allocate more cognitive effort to individual relationships and thus have smaller ego networks on average. All in all, even though structurally ENMs are known to be similar in both offline and online social networks, our results indicate that relationships on Twitter tend to nurture more negativity than offline contexts.

Enhancing the Decision Tree Algorithm to Improve Performance Across Various Datasets

Background: The Village Fund is an initiative by the central government to promote equitable regional development. However, it has also led to corruption. Many Indonesians share their opinions on the Village Fund on social media platforms like X, and news coverage is extensive on portals like detik.com. Objective: This study aims to classify data from social media and news coverage to enhance understanding. Methods: The research improves the decision tree algorithm by integrating other algorithms and techniques such as XGBoost and SMOTE. Ensuring high accuracy is vital for the credibility of machine learning classifications among the public. The study uses two different datasets, necessitating varied testing approaches. For the news portal dataset, a single test with seven labels is conducted, followed by enhancement with XGBoost. The X dataset undergoes two tests with datasets of 1200 and 3078 entries, using three labels. Conclusion: The evaluation results indicate that the highest accuracy achieved with the news portal data was 82%, thanks to a combination of decision tree algorithms with various parameters and the balancing effect of SMOTE. For the Twitter dataset with 3078 entries, the highest accuracy reached 95%, attributed to the application of ensemble techniques, particularly boosting.

CommentClass: A Robust Ensemble Machine Learning Model for Comment Classification

Enormous amounts of data are generated in the form of feedback or comments from online platforms such as social media, e-commerce, education, and programming. This feedback and comments hold significant value for making important strategic decisions; therefore, effectively analyzing them poses a major challenge. This research addresses the imperative need for an efficient comment classification model. To fill this research gap, we propose a robust ensemble machine learning (ML) model called CommentClass (RF+AdaBoost+SVM+Soft-Voting), specifically designed for the comment classification task. First, we developed eight (08) pipelines using various combinations of ML algorithms. Next, the fundamental ensemble techniques such as stacking, blending, hard-voting, soft-voting, and averaging are incorporated into these pipelines to improve comment classification performance. These ensemble models are able to discern the latent characteristics of diverse text comments, classifying them to achieve superior accuracy. The proposed CommentClass ensemble model achieved an impressive accuracy and F1-score of approximately 98% for comment classification on the YouTube dataset. This result represents an improvement in accuracy by approximately +3% compared to prior research on the same dataset. Moreover, the proposed CommentClass model obtained higher F1-scores of 90.26%, 87.04%, and 75.74%, on the Spambase, IMDB, and Twitter datasets, respectively, compared to other sophisticated models. Furthermore, the proposed CommentClass model exhibited significant accuracy on the SMS dataset and two distinct synthetic datasets.

TCKGCN: Graph convolutional network for aspect-based sentiment analysis with three-channel knowledge fusion

At present, most aspect-based sentiment analysis studies addresses the syntactic dependency problem of review statements by building syntactic dependency graphs. However, this approach is insensitive to syntactic structural information and susceptible to noise interference. Existing fusion methods are unable to effectively capture the complementary relationships between information, leading to large and less efficient model structures. To tackle these issues, we introduce a unique and original scheme that fuses semantic information, ConceptNet conceptual knowledge, and SenticNet affective knowledge, intending to improve the capability of the model to capture sentence features. Specifically, the process involves four steps: first, a semantic attention mechanism is constructed as the first channel to parse the semantic relationship within the context; second, the expressive meaning of the aspects is enhanced with concept mapping as the second channel, with concepts injected into aspectual words; third, the representation ability of sentence dependency graphs is strengthened through adding the affective knowledge of SenticNet as a third channel. Subsequently, the sensitivity of the model to affective information is increased, and finally, the optimization of aspect and context coordination is strengthened by integrating an interactive attention mechanism. The overall findings underscore the efficacy of the model, showing accuracy improvements of 1.05 % and 2.80 % on the Twitter and Lap14 datasets, respectively, compared to state-of-the-art models. Moreover the macro-F1 scores showed significant improvements of 1.99 % and 1.61 %, respectively, enabling more effective capturing of aspect-specific sentiment expressions.

Rumor gatekeepers: Unsupervised ranking of Arabic twitter authorities for information verification

The advent of online social networks (OSNs) has catalyzed the formation of novel learning communities. Identifying experts within OSNs has become a critical component for facilitating knowledge exchange and enhancing self-awareness, particularly in contexts such as rumor verification processes. Research efforts aimed at locating authorities in OSNs are scant, largely due to the scarcity of annotated datasets. This work represents a contribution to the domain of unsupervised learning to address the challenge of authorities’ identification in Twitter. We have employed advanced natural language processing technique to transfer knowledge concerning topics in the Arabic language and to discern the semantic connections among candidates within Twitter in zero-shot learning. We take advantage of the Single-labeled Arabic News Articles Dataset (SANAD) to perform the process of extracting domain features and applying these features in finding authorities using the Authority Finding in Arabic Twitter (AuFIN) dataset. Our evaluation assessed the extent of extracted topical features transferred and the efficacy of authorities’ retrieval in comparison to the latest unsupervised models in this domain. Our approach successfully extracted and integrated the limited available topical semantic features of the language into the representation of candidates. The findings indicate that our hybrid model surpasses those that rely solely on lexical features of language and network topology, as well as other contemporary approaches to topic-specific expert finding.

Psychological Stress Detection Using Transformer-Based Models

Stress is a significant mental health problem that results in a lack of concentration. It has been more widely identified through social media since people who are under stress usually post about their physical pain and tiredness. However, stress assessment through social media by professionals can be expensive and time-consuming. The research aimed to produce a stress detection system trained using a Twitter dataset to predict stress using the user’s input sentence. The experiments that were done in the research used transformer-based models such as Bidirectional Encoder Representations from Transformers (BERT) and Robustly Optimized BERT (RoBERTa). The research involved data pre-processing, model training, and model evaluation to ensure high-quality train data. Since the data were imbalanced, data trimming was performed in pre-processing to select data randomly until the balance matched. This process ensured the model’s effectiveness in the training and evaluation stages. The features used in these experiments were features from each pre-trained model. In evaluating the model, accuracy, loss, and F1 score were used as metrics. In the result, for BERT, accuracy reaches 0.848 with an F1 score of 0.847. Meanwhile, RoBERTa has an accuracy of 0.837 and 0.834. The results prove that BERT and RoBERTa can be used to classify stress with accuracy and an F1 score above 0.8. The experiment result shows that the BERT deep learning model can detect stress using the Twitter datasets.

Data oversampling and imbalanced datasets: an investigation of performance for machine learning and feature engineering

The classification of imbalanced datasets is a prominent task in text mining and machine learning. The number of samples in each class is not uniformly distributed; one class contains a large number of samples while the other has a small number. Overfitting of the model occurs as a result of imbalanced datasets, resulting in poor performance. In this study, we compare different oversampling techniques like synthetic minority oversampling technique (SMOTE), support vector machine SMOTE (SVM-SMOTE), Border-line SMOTE, K-means SMOTE, and adaptive synthetic (ADASYN) oversampling to address the issue of imbalanced datasets and enhance the performance of machine learning models. Preprocessing significantly enhances the quality of input data by reducing noise, redundant data, and unnecessary data. This enables the machines to identify crucial patterns that facilitate the extraction of significant and pertinent information from the preprocessed data. This study preprocesses the data using various top-level preprocessing steps. Furthermore, two imbalanced Twitter datasets are used to compare the performance of oversampling techniques with six machine learning models including random forest (RF), SVM, K-nearest neighbor (KNN), AdaBoost (ADA), logistic regression (LR), and decision tree (DT). In addition, the bag of words (BoW) and term frequency and inverse document frequency (TF-IDF) features extraction approaches are used to extract features from the tweets. The experiments indicate that SMOTE and ADASYN perform much better than other techniques thus providing higher accuracy. Additionally, overall results show that SVM with ’linear’ kernel tends to attain the highest accuracy and recall score of 99.67% and 1.00% on ADASYN oversampled datasets and 99.57% accuracy on SMOTE oversampled dataset with TF-IDF features. The SVM model using 10-fold cross-validation experiments achieved 97.40 mean accuracy with a 0.008 standard deviation. Our approach achieved 2.62% greater accuracy as compared to other current methods.

MFAE: Multimodal Fusion and Alignment for Entity-level Disinformation Detection

Nowadays, the dissemination of disinformation on social media has evolved from a purely textual form to multiple modalities consisting of both text and images. This further amplifies the misleading and deceptive nature of disinformation. Overcoming the misleading and confusing noise to achieve accurate disinformation detection presents a significant challenge. To address this challenge, we propose a method named Multimodal Fusion and Alignment for Entity-level Disinformation Detection (MFAE). MFAE first uses an improved dynamic routing algorithm to extract more comprehensive semantic visual entity features. Then, a graph matching network is used to capture the correspondences between entities within modalities. The experiment shows that MFAE is capable of capturing textual and visual semantic information more comprehensively. On the TWITTER and WEIBO datasets, MFAE achieves accuracy improvements of approximately 2.0% and 7.5%, respectively, compared to the state-of-the-art methods, resulting in accuracy of 89.5% and 96.7%.

Using social media metrics and linked survey data to understand survey behaviors

Introduction & BackgroundLinking social media and survey data at the individual level has the potential to add evidence to a variety of research questions. To make this data openly available to others, social media data need to be converted into useful metrics that minimise issues of disclosure while maximising utility. This research explores linkages of Twitter data and survey data in the Understanding Society Innovation panel, focusing on the usage of non-disclosive metrics created from Twitter data alongside the similarly anonymised survey data. Objectives & ApproachThe Innovation Panel asked for consent to link Twitter data to survey responses and data has been collected from the Twitter API. However, Twitter’s unstructured nature necessitates creating measures that can be used jointly with linked survey data. We have developed a framework to create social media metrics that can be combined with survey data that also remove any disclosive data, so these data can be widely shared for maximum utility. The current research analyses these data to understand what the metrics look like through presentation of descriptive statistics. We also begin to show these data may be used in combination with survey data through inclusion of a set of metrics in logistic regression models predicting attrition and measurement of mental health. Relevance to Digital FootprintsSocial media is a prevalent aspect of social life and leaves a substantial digital footprint. However, there are a number of limitations to these data, including a lack of understanding of who is producing the data, and having the ability to relate these to a variety of specific (and possibly higher quality) measures for a representative sample of the population. Linkage to surveys address these problems and can lead to new research opportunities using digital footprint data. ResultsWhile small sample sizes impact the power of some analyses, the methods developed are illustrative of ways to use this novel data source. Results show that there is high variation in the created metrics, and initial analysis shows that the inclusion of a set of user-level Twitter data is not significantly related to attrition. However, more accounts followed on Twitter and the number of user retweets are significantly related to higher levels of mental distress on the GHQ scale. Conclusions & ImplicationsOverall, there is some evidence that social media helps to understand survey outcomes, perhaps more so on measurement outcomes. This study provides an initial start on how to use these curated linked social media and survey data, and we note there are other social media networks that we can apply this strategy to; for example, LinkedIn, particularly with changes made to Twitter (X).