Figure 2 - uploaded by Abu Nowshed Chy
Content may be subject to copyright.
Source publication
Stance detection in twitter aims at mining user stances expressed in a tweet towards a single or multiple target entities. To tackle this problem, most of the prior studies have been explored the traditional deep learning models, e.g., LSTM and GRU. However, in compared to these traditional approaches, recently proposed densely connected Bi-LSTM an...
Context in source publication
Context 1
... this section, we describe the details of our proposed neural ensemble model (PNEM) for twitter stance detection. Figure 2 depicts an overview of our proposed framework. At first, we utilize the multi-kernel convolution filters to extract higher-level feature sequences from the target appended tweet embeddings. ...
Similar publications
Social media enable their users to be connected with a diverse group of people increasing their chances of coming across divergent viewpoints. Thus, network diversity is a key issue for understanding the potentials of social media for creating a cross-cutting communication space that is one of the premises of a functioning democracy. This article a...
Die zunehmende gesellschaftliche Nutzung sozialer Medien als öffentliches Kommunikationsmedium erzeugt eine kontinuierlich wachsende Datenmenge nutzergenerierter Inhalte. In Krisen und Katastrophen, aber auch während Großveranstaltungen können lagerelevante Informationen aus sozialen Medien gewonnen und in ein Lagebild integriert und visualisiert w...
Citations
... The performance of stance detection models employing supervised learning in recent studies slightly varies. Some approaches employing traditional machine learning models such as Support Vector Machine (SVM) achieved F1 scores of 69% (Mohammad, Sobhani, and Kiritchenko 2017) and 63.6% (Elfardy and Diab 2016), while the work leveraging a bidirectional LSTM with a fast-text embedding layer reported an F1 score of 72.1% (Siddiqua, Chy, and Aono 2019). However, recently, several studies approach stance detection by using BERT-based models (Kawintiranon and Singh 2021;Liu et al. 2021;Alturayeif, Luqman, and Ahmed 2022;Glandt et al. 2021;Clark et al. 2021;Barbieri et al. 2020) reporting average F1 scores in range approximately between 0.7 to 0.9. ...
Abortion remains one of the most controversial topics, especially after overturning Roe v. Wade ruling in the United States. Previous literature showed that the illegality of abortion could have serious consequences, as women might seek unsafe pregnancy terminations leading to increased maternal mortality rates and negative effects on their reproductive health. Therefore, the stances of the abortion-related Facebook posts were analyzed at the state level in the United States from May 4 until June 30, 2022, right after the Supreme Court’s decision was disclosed. In more detail, a pre-trained Transformer architecture-based model was fine-tuned on a manually labeled training set to obtain a stance detection model suitable for the collected dataset. Afterward, we employed appropriate statistical tests to examine the relationships between public opinion regarding abortion, abortion legality, political leaning, and factors measuring the overall population’s health, health knowledge, and vulnerability per state. We found that infant mortality rate, political affiliation, abortion rates, and abortion legality are associated with stances toward abortion at the state level in the US. While aligned with existing literature, these findings indicate how public opinion, laws, and women’s and infants’ health are related, as well as how these relationships can be demonstrated by using social media data.
... RNNs are good at processing sequential data. In Twitter stance detection, Siddiqua proposed a variant that utilizes bi-LSTMs and nested LSTMs to capture long-term dependencies, where each module is enhanced with an attention mechanism [13]. In the two-target stance detection task, Liu et al. used multiple LSTM layers to encode target-related regions [14]. ...
Stance information has a significant influence on market strategy, government policy, and public opinion. Users differ not only in their polarity but also in the degree to which they take a stand. The traditional classification of stances is quite simple and cannot fully depict the diversity of stances. At the same time, traditional approaches ignore user sentiment features when expressing their stances. As a result, this paper develops a multi-stance detection model by fusing sentiment features. First, a five-category stance indicator system is built based on the LDA model, then sentiment features are extracted from the reviews using the sentiment lexicon, and finally, stance detection is implemented using a hybrid neural network model. The experiment shows that the proposed method can classify stances into five categories and perform stance detection more accurately.
... Prior work on stance detection investigated several approaches such as designing specific features [5,6], building deep learning architectures [7,8], and utilizing social network [9]. As the target plays a crucial role in stance detection methods, numerous researchers have also sought to develop models for topics not encountered * Correspondence: mucahidkutlu@gmail.com ...
... The studies focusing on social media accounts also utilized social network-based features such as retweeted accounts [23] and followers [9]. A number of studies also employed deep learning models such as bi-LSTM [8] and bi-GRU [30]. Ghosh et al. [7] compare several stance detection methods and report that fine-tuned BERT model has the highest performance on the SemEval 2016 dataset. ...
... Other methods, such as topic modeling [26] and morality analysis [22,23] have been employed to identify underlying themes and reasoning behind a stance. BERT and other deep learning methods have been leveraged to achieve higher accuracy in stance detection [25]. ...
In today’s social media landscape, personal opinions on controversial topics are widespread. While some platforms provide structured environments for discussing such matters, fostering cross-cutting communication among individuals, understanding how people engage in these discussions remains a challenge. This study aims to understand the dynamics of discussing controversial topics, focusing specifically on the topic of abortion. Using an aspect-based approach, we employ BERT-based topic modeling and attention mechanisms to identify key aspects of debates. Through clustering, we identify highly polarizing aspects and examine the contextual nuances and sentiment surrounding them. Our methodology enhances our understanding of cross-cutting communication on controversial topics and offers an in-depth analysis of consensus and disagreement among participants. Our study contributes to the field of stance analysis, revealing opportunities for mutual understanding and uncovering diverse perspectives on controversial issues. (Warning: this paper contains content that may be triggering.)
... The deep learning techniques used in their approach are Convolutional Neural Networks (CNN) and Recurrent Convolutional Neural Networks (RCNN). An attention based neural ensemble model is proposed for tweet stance detection in [27]. The authors used Bi-LSTM and nested LSTM architectures in the ensemble. ...
Detecting informative tweets is very important to the government or non-government organizations during a disaster. Most of the literature works focused on either text or image separately for getting informative tweets. A very few existing works used multi-modal information such as both image and text to identify the informative tweets. However, the existing works do not give much performance on multi-modal informative tweets. There is a chance to lose useful information in critical times. Hence, we propose a novel approach to identify the multi-modal informative tweets during a disaster. Our proposed method comprises the pre-trained RoBERTa and VGG-16 models to extract the text and image features, respectively. The outputs of these two models are combined using a multiplicative fusion technique. Experiments are conducted on diverse disaster datasets such as Hurricane Maria, Hurricane Harvey, California wildfires, Iraq-Iran earthquake, Hurricane Irma, and Mexico earthquake. Experimental results demonstrated that the proposed method outperforms the existing baseline methods on various parameters.
... Popular tweet classification tasks include stance detection [36], sentiment analysis [37] emotion recognition [35], hate speech detection [4], and so on. Several studies explored SVM-based approaches [11,42], CNN-based models [47,52] and RNN-based models [43,58], respectively. Later, with the prevalence of Transformer architecture [45] and pre-trained language models [10,27], BERT-based approaches [15,25] attracted a lot of attention and had proven their effectiveness in this task. ...
Social media classification tasks (e.g., tweet sentiment analysis, tweet stance detection) are challenging because social media posts are typically short, informal, and ambiguous. Thus, training on tweets is challenging and demands large-scale human-annotated labels, which are time-consuming and costly to obtain. In this paper, we find that providing hashtags to social media tweets can help alleviate this issue because hashtags can enrich short and ambiguous tweets in terms of various information, such as topic, sentiment, and stance. This motivates us to propose a novel Hashtag-guided Tweet Classification model (HashTation), which automatically generates meaningful hashtags for the input tweet to provide useful auxiliary signals for tweet classification. To generate high-quality and insightful hashtags, our hashtag generation model retrieves and encodes the post-level and entity-level information across the whole corpus. Experiments show that HashTation achieves significant improvements on seven low-resource tweet classification tasks, in which only a limited amount of training data is provided, showing that automatically enriching tweets with model-generated hashtags could significantly reduce the demand for large-scale human-labeled data. Further analysis demonstrates that HashTation is able to generate high-quality hashtags that are consistent with the tweets and their labels. The code is available at https://github.com/shizhediao/HashTation.
... Previous stance identification on Twitter was cast either as (1) a classification problem, learning to predict the stance value of a tweet towards a given target; or (2) an inference problem, when a tweet may entail, contradict, or does not imply the target. Classification methods: Neural architectures, relying on RNNs, CNNs, and/or attention mechanisms, dominate these methods (Augenstein et al., 2016;Du et al., 2017;Sun et al., 2018;Siddiqua et al., 2019). BERT was later shown to be the best model overall for stance detection on the SemEval2016 Task 6 (Ghosh et al., 2019). ...
... Previous works for stance detection mainly focus on the in-target setting (Mohammad et al., 2016b;Du et al., 2017;Siddiqua et al., 2019;Caragea, 2019, 2021a) where the test target has always been seen in the training stage. Recently, cross-target stance detection (Augenstein et al., 2016;Xu et al., 2018;Liang et al., 2021) and zero-shot stance detection (Allaway and McKeown, 2020;Allaway et al., 2021;Liu et al., 2021a;Liang et al., 2022a,b;Li et al., 2023) have also attracted a lot of attention. ...
... From the user's friends and followers, additional knowledge is extracted from the author, forwarding, quoting, and replying. Siddiqua et al. [28] proposed a neural ensemble model that uses the advantages of Bi-LSTM variants to better learn long-term dependencies, in which each module with an attention mechanism is combined to amplify the contribution of essential elements in the final representation. ...
In the age of social networks, the number of tweets sent by users has led to a sharp rise inpublic opinion. Public opinions are closely related to user stances. User stance detection has becomean important task in the field of public opinion. However, previous studies have not distinguishedbetween user viewpoints and stances. These studies usually detected stance from the perspective ofthe tweet level but rarely the user level. Therefore, in this paper, we defined user stance, which is theuser viewpoint (support, oppose, and neutral) toward the entire target event process. On this basis,we put forward a user stance detection method based on external commonsense knowledge (suchas SenticNet) and environment information (such as a user’s historical tweets, topic information,and neighbor tweets) and denote this method as ECKEI. First, in order to better integrate externalcommonsense knowledge into the neural network, we improved BiLSTM and called it CK-BiLSTM forcomplementary commonsense information to the memory cell. Secondly, we used LDA to extract thetopic of user tweets and designed a topic-driven module to capture the users’ neighbors’ information.Finally, we used the attention mechanism to integrate information from users’ historical tweets andneighbors’ tweets obtained through topic information; then, we used the softmax layer to classify userstances into the support, neutral and oppose classes. In this paper, we conducted experiments andassessments on datasets containing information on Brexit and the elections to verify the practicabilityand effectiveness of our proposed method. Extensive experimental results on the Brexit and electionsdatasets show that our approach outperforms six baseline methods (SVM-ngram, NB, MTTRE (RNN),Pkudblab (CNN), TAAT, and Aff-feature). We use the average micro-F1 and average accuracy tomeasure performance on the detection of a user’s stance. The ECKEI model makes improvements of4.30–16.89% and 1.22–16.58% on the Brexit and election datasets, respectively, in terms of averagemicro-F1. Meanwhile, this model makes improvements of 4.24–17.46% and 0.48–14.64% on the Brexitand election datasets, respectively, in terms of average accuracy. Our model makes improvements of5.34–17.30% and 2.65–19.73%, respectively, on the Brexit and election datasets in terms of averagerecall.
... Agricultural reforms or knowledge is a process of increasing pay and socioeconomic strength to speed up the fulfilment of basic needs and managements while also constructing a range of benefits that will reduce family unit vulnerability to physical, communal, and economic shocks. This sector is extremely important to small-scale farmers (Bamwesigye, 2008;Siddiqua et al., 2019). ...
Global climate change is a change in the long-term weather patterns that characterize the regions of the world. So, if we look at the situation in Pakistan it is affecting rapidly that’s why Production and growth rate of Agriculture is badly affected because of it. Its destruction on Agriculture sector is massive in Pakistan. Changing in climate is directly affecting the livelihoods of Agricultural communities in Pakistan. It has also become the reason of rising temperature and lot of other problems in agriculture sector like food production, quality, and quantity of food is decreasing. Rising temperature effecting yields a lot in sense of production. The worst and biggest effect of climate change is flood. Around 3.4 million hector area of agriculture damaged by flood during 2010 and2011 and now in 2022 same as previously 9.1 million acres agricultural land is affected by floods and now the flood situation is worse than the past. Being an agrarian country more than 70% labour force of Pakistan depends on the agriculture. Due to climate changes people living or depending on agricultural sector for their livelihoods are facing much serious problem as their production in agricultural sector is much less than before. People of Pakistan facing so many challenges regarding Livelihood because they are depending on the output of their farms which is affecting severely. This study was conducted in district Muzaffargarh, Punjab, Pakistan.
