Multi-label emotion classification in texts using transfer learning

Ameer, Iqra; Bölücü, Necva; Siddiqui, Muhammad Hammad Fahim; Can, Burcu; Sidorov, Grigori; Gelbukh, Alexander

doi:10.1016/j.eswa.2022.118534

Cited by 55 publications

(30 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A few prior works tried to understand these fine-grained reasons for vaccine hesitancy, mostly by manual analysis (Bonnevie et al 2020) Multi-label Classification: Multi-label classification is a long-studied problem, and several approaches have been applied in various sub-domains of social media analysis, such as emotion detection from tweets (Mukherjee et al 2021;Ameer et al 2023), disaster mitigation (Chowdhury et al 2020) and symptom detection (Jarynowski et al 2021). Looking out to general domains, entailment-based methods (Wang et al 2021) and generative models (Simig et al 2022) (which we apply in this work) have been applied to classification tasks.…”

Section: Related Workmentioning

confidence: 99%

Winds of Change: Impact of COVID-19 on Vaccine-Related Opinions of Twitter Users

Poddar

Mondal

Misra³

et al. 2022

ICWSM

View full text Add to dashboard Cite

Today, administering COVID-19 vaccines at a societal scale has been deemed as the most appropriate way to defend against the COVID-19 pandemic. This global vaccination drive naturally fueled a possibility of pro-vax and anti-vax users strongly expressing their supports and concerns regarding the vaccines in online social media platforms. Understanding this online discourse is crucial for policy makers. This understanding is likely to impact the success of vaccination drives and might even impact the final outcome of our fight against the pandemic. The goal of this work is to improve this understanding using the lens of Twitter-discourse data. We first develop a classifier to categorize users according to their vaccine-related stance with high precision (97%). Using this method we detect and investigate specific user-groups who posted about vaccines in pre-COVID and COVID times. We specifically identify distinct topics that these users talk about, and investigate how vaccine-related discourse has changed between pre-COVID times and COVID times. Finally, for the first time, we investigate the change of vaccine-related stances in Twitter users and shed light on potential reasons for such changes in stance.

show abstract

Section: Related Workmentioning

confidence: 99%

Winds of Change: Impact of COVID-19 on Vaccine-Related Opinions of Twitter Users

Poddar

Mondal

Misra³

et al. 2022

ICWSM

View full text Add to dashboard Cite

show abstract

“…To evaluate VTI approaches, we apply the evaluation metrics which are widely used in multi-label classification studies [19]. For each function 𝑓 𝑖 in the test set, the types of 𝑓 𝑖 is represented by a vector 𝑌 𝑖 such that for 𝑗 ∈ [1, ‖𝑇 ‖], 𝑌 𝑖𝑗 = 1 if 𝑇 𝑗 ∈ 𝑇 is one of the vulnerability types of 𝑓 , otherwise 𝑌 𝑖𝑗 = 0.…”

Section: Evaluation Metricsmentioning

confidence: 99%

“…In this work, for both deep learning models, we follow the typical architectures of classifiers for the general multilabel text classification task instead of designing a complex model. We evaluate the performance of the multi-label classification model using Word2vec proposed by [19]. They use Word2vec to construct a word embedding layer followed by two Bi-LSTM layers, an attention layer, a fully connected, and the sigmoid activation function (Fig.…”

Section: Deep Learning Approachesmentioning

confidence: 99%

“…The model also feeds the embedding to a fully connected layer and either the sigmoid function to compute the classification. For both models, we use binary cross entropy [19] to compute the loss between the classification and the ground truth.…”

Section: Deep Learning Approachesmentioning

confidence: 99%

“…In this work, we make the first step to explore the problem of VTI by using the existing techniques in software engineering (SE) and natural language processing (NLP). Particularly, we model VTI as the multi-label text classification task in NLP [18,19,20]. This is reasonable because each vulnerable function 𝑓 could be considered as a document 𝑑, and the type set of the vulnerabilities in 𝑓 could be the label set of 𝑑.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Can an old fashioned feature extraction and a light-weight model improve vulnerability type identification performance?

Vo¹,

Nguyen²

2023

Information and Software Technology

View full text Add to dashboard Cite

A Bayesian transfer sparse identification method for nonlinear ARX systems

Zhang,

Luan,

Ding

et al. 2024

Adaptive Control & Signal

View full text Add to dashboard Cite

SummaryIn this paper, we design a transfer sparse identification algorithm under the Bayesian framework through introducing other system knowledge into the system to be identified. This method provides a new identification solution for a nonlinear autoregressive model with exogenous inputs (NARX). The estimates of the transferred parameters are calculated by adding the transfer correction term to the un‐transferred estimates. To achieve this, a joint prior distribution is devised for the parameters, ultimately enhancing the efficient utilization of existing data, reducing the reliance on new data, and achieving more accurate identification. The maximized marginal likelihood method is used to find the transfer gain and the transfer information matrix in the transfer correction term. Meanwhile, in order to make the algorithm automatically adapt to different data, we design an automatic structure detection method based on the transfer framework. The method automatically determines the sparsity threshold based on the maximum inter‐class variance. Two examples are provided to demonstrate the advantages of our algorithm.

show abstract

Multi-label emotion classification in texts using transfer learning

Cited by 55 publications

References 28 publications

Winds of Change: Impact of COVID-19 on Vaccine-Related Opinions of Twitter Users

Winds of Change: Impact of COVID-19 on Vaccine-Related Opinions of Twitter Users

Can an old fashioned feature extraction and a light-weight model improve vulnerability type identification performance?

A Bayesian transfer sparse identification method for nonlinear ARX systems

Contact Info

Product

Resources

About