%0 Journal Article %1 9126211 %A Dong, H. %A Wang, W. %A Huang, K. %A Coenen, F. %D 2020 %J IEEE Transactions on Neural Networks and Learning Systems %K attention_mechanisms automated_social_tagging bibsonomy citeulike deep_learning folksonomy guided-attention jman knowledge multi-label_classification multilabel myown ontologies recurrent_neural_networks rnns semantic_relations semantics social_annotation social_tagging tagging zhihu %P 1-15 %R 10.1109/TNNLS.2020.3002798 %T Automated Social Text Annotation With Joint Multilabel Attention Networks %U https://ieeexplore.ieee.org/document/9126211 %X Automated social text annotation is the task of suggesting a set of tags for shared documents on social media platforms. The automated annotation process can reduce users' cognitive overhead in tagging and improve tag management for better search, browsing, and recommendation of documents. It can be formulated as a multilabel classification problem. We propose a novel deep learning-based method for this problem and design an attention-based neural network with semantic-based regularization, which can mimic users' reading and annotation behavior to formulate better document representation, leveraging the semantic relations among labels. The network separately models the title and the content of each document and injects an explicit, title-guided attention mechanism into each sentence. To exploit the correlation among labels, we propose two semantic-based loss regularizers, i.e., similarity and subsumption, which enforce the output of the network to conform to label semantics. The model with the semantic-based loss regularizers is referred to as the joint multilabel attention network (JMAN). We conducted a comprehensive evaluation study and compared JMAN to the state-of-the-art baseline models, using four large, real-world social media data sets. In terms of F₁, JMAN significantly outperformed bidirectional gated recurrent unit (Bi-GRU) relatively by around 12.8%-78.6% and the hierarchical attention network (HAN) by around 3.9%-23.8%. The JMAN model demonstrates advantages in convergence and training speed. Further improvement of performance was observed against latent Dirichlet allocation (LDA) and support vector machine (SVM). When applying the semantic-based loss regularizers, the performance of HAN and Bi-GRU in terms of F₁ was also boosted. It is also found that dynamic update of the label semantic matrices (JMANd) has the potential to further improve the performance of JMAN but at the cost of substantial memory and warrants further study.

@article{9126211, abstract = {Automated social text annotation is the task of suggesting a set of tags for shared documents on social media platforms. The automated annotation process can reduce users' cognitive overhead in tagging and improve tag management for better search, browsing, and recommendation of documents. It can be formulated as a multilabel classification problem. We propose a novel deep learning-based method for this problem and design an attention-based neural network with semantic-based regularization, which can mimic users' reading and annotation behavior to formulate better document representation, leveraging the semantic relations among labels. The network separately models the title and the content of each document and injects an explicit, title-guided attention mechanism into each sentence. To exploit the correlation among labels, we propose two semantic-based loss regularizers, i.e., similarity and subsumption, which enforce the output of the network to conform to label semantics. The model with the semantic-based loss regularizers is referred to as the joint multilabel attention network (JMAN). We conducted a comprehensive evaluation study and compared JMAN to the state-of-the-art baseline models, using four large, real-world social media data sets. In terms of F₁, JMAN significantly outperformed bidirectional gated recurrent unit (Bi-GRU) relatively by around 12.8%-78.6% and the hierarchical attention network (HAN) by around 3.9%-23.8%. The JMAN model demonstrates advantages in convergence and training speed. Further improvement of performance was observed against latent Dirichlet allocation (LDA) and support vector machine (SVM). When applying the semantic-based loss regularizers, the performance of HAN and Bi-GRU in terms of F₁ was also boosted. It is also found that dynamic update of the label semantic matrices (JMANd) has the potential to further improve the performance of JMAN but at the cost of substantial memory and warrants further study.}, added-at = {2020-06-25T22:43:32.000+0200}, author = {{Dong}, H. and {Wang}, W. and {Huang}, K. and {Coenen}, F.}, biburl = {https://www.bibsonomy.org/bibtex/22fe9a55ff296ec72758f582f86f4b94b/hangdong}, description = {Automated Social Text Annotation With Joint Multilabel Attention Networks - IEEE Journals & Magazine}, doi = {10.1109/TNNLS.2020.3002798}, interhash = {787418d62f4d745e9ca50ae82a799564}, intrahash = {2fe9a55ff296ec72758f582f86f4b94b}, issn = {2162-2388}, journal = {IEEE Transactions on Neural Networks and Learning Systems}, keywords = {attention_mechanisms automated_social_tagging bibsonomy citeulike deep_learning folksonomy guided-attention jman knowledge multi-label_classification multilabel myown ontologies recurrent_neural_networks rnns semantic_relations semantics social_annotation social_tagging tagging zhihu}, pages = {1-15}, timestamp = {2020-06-25T22:43:32.000+0200}, title = {Automated Social Text Annotation With Joint Multilabel Attention Networks}, url = {https://ieeexplore.ieee.org/document/9126211}, year = 2020 }