To detect multiple coexisting emotions from public emergency opinions, this article proposes a novel two-stage multiple coexisting emotion-detection model. First, the text semantic feature extracted through bidirectional encoder representation from transformers (BERT) and the emotion lexicon feature extracted through the emotion dictionary are fused. Then, the emotion subjectivity judgement and multiple coexisting emotion detection are performed in two separate stages. In the first stage, we introduce synthetic minority oversampling technique (SMOTE) to enhance the balance of data distribution and select the optimal classifier to recognise opinion texts with emotion. In the second stage, the label powerset (LP)-SMOTE is proposed to increase the number of the minority category samples, and multichannel emotion classifiers and the decision mechanism are employed to recognise different types of emotions and determine the final coexisting emotion labels. Finally, the Weibo data about coronavirus disease 2019 (COVID-19) are collected to verify the effectiveness of the proposed model. Experiment results indicate that the proposed model outperforms state-of-the-art models, with the F1_macro of 0.8532, the F1_micro of 0.8333, and the hamming loss of 0.0476. The emotion detection results are conducive to decision-making for public emergency departments.
ShiWXueGCHeSY. Literature review of network public opinion research from the perspective of sentiment. Doc Inf Knowl2022; 39(1): 105–118.
2.
ZhuYWuO. Elementary discourse units with sparse attention for multi-label emotion classification. Knowl-Based Syst2022; 240: 108114.
3.
ChandraRKrishnaA. COVID-19 sentiment analysis via deep learning during the rise of novel cases. PLoS ONE2021; 16(8): e0255615.
4.
Cabrera-DiegoLABessisNKorkontzelosI. Classifying emotions in stack overflow and JIRA using a multi-label approach. Knowl-Based Syst2020; 195: 105633.
5.
ReadJPfahringerBHolmesG, et al. Classifier chains for multi-label classification. Mach Learn2011; 85: 333–359.
6.
ShanJHouCTaoH, et al. Co-learning binary classifiers for LP-based multi-label classification. Cognit Syst Res2019; 55: 146–152.
7.
ZhangMLZhouZH. A review on multi-label learning algorithms. IEEE Trans Knowl Data Eng2014; 26(8): 1819–1837.
8.
ClareAKingRD. Knowledge discovery in multi-label phenotype data. In: De RaedtLSiebesA (eds) Principles of data mining and knowledge discovery: 5th European conference, PKDD 2001. Berlin: Springer, 2001, pp. 42–53.
9.
ElisseeffAWestonJ. A kernel method for multi-labelled classification. Adv Neural Inf Process Syst2001; 14: 681–687.
10.
PapanikolaouYTsoumakasGKatakisI. Hierarchical partitioning of the output space in multi-label data. Data Knowl Eng2018; 116: 42–60.
11.
YangPSunXLiW, et al. SGM: sequence generation model for multi-label classification. arXiv preprint arXiv1806.04822, 2018.
12.
HeHXiaR. Joint binary neural network for multi-label learning with applications to emotion classification. In: Natural language processing and Chinese computing: 7th CCF international conference, NLPCC 2018, Hohhot, China, 26–30 August 2018, proceedings, part I 7, pp. 250–259. Cham: Springer International Publishing.
13.
ChenGYeDXingZ, et al. Ensemble application of convolutional and recurrent neural networks for multi-label text categorization. In: 2017 International joint conference on neural networks (IJCNN), Anchorage, AK, 14–19 May 2017, pp. 2377–2383. New York: IEEE.
14.
DevlinJChangMWLeeK, et al. Bert: pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv 1810.04805, 2018.
15.
HuJKangXNishideS, et al. Text multi-label sentiment analysis based on Bi-LSTM. In: 2019 IEEE 6th international conference on cloud computing and intelligence systems (CCIS), Singapore, 19–21 December 2019, pp. 16–20. New York: IEEE.
16.
BendjoudiIVanderhaegenFHamadD, et al. Multi-label, multi-task CNN approach for context-based emotion recognition. Inf Fusion2021; 76: 422–428.
17.
XieTWeiYYChenWF, et al. Parallel evolution and response decision method for public sentiment based on system dynamics. Eur J Oper Res2020; 287(3): 1131–1148.
18.
LiuXZhengLJiaX, et al. Public opinion analysis on novel coronavirus pneumonia and interaction with event evolution in real world. IEEE Trans Comput Soc Syst2021; 8(4): 1042–1051.
19.
NasukawaTYiJ. Sentiment analysis: capturing favorability using natural language processing. In: Proceedings of the 2nd international conference on knowledge capture, Sanibel Island, FL, 23–25 October 2003, pp. 70–77. New York: Association for Computing Machinery.
RevathyGAlghamdiSAAlahmariSM, et al. Sentiment analysis using machine learning: progress in the machine intelligence for data science. Sustain Energy Technol Assess2022; 53: 102557.
22.
AlBadaniBShiRDongJ. A novel machine learning approach for sentiment analysis on Twitter incorporating the universal language model fine-tuning and SVM. Appl Syst Innov2022; 5(1): 13.
23.
BhuvaneshwariPRaoANRobinsonYH, et al. Sentiment analysis for user reviews using Bi-LSTM self-attention based CNN model. Multimed Tools Appl2022; 81(9): 12405–12419.
24.
KhanLAmjadAAfaqKM, et al. Deep sentiment analysis using CNN-LSTM architecture of English and Roman Urdu text shared in social media. Appl Sci2022; 12(5): 2694.
25.
HuangBZhangJJuJ, et al. CRF-GCN: an effective syntactic dependency model for aspect-level sentiment analysis. Knowl-Based Syst2023; 260: 110125.
26.
WangTYangW. Review of text sentiment analysis methods. Comput Eng Appl2021; 57(12): 11–24.
PengDZhouMLiuC, et al. Human–machine dialogue modelling with the fusion of word-and sentence-level emotions. Knowl-Based Syst2020; 192: 105319.
29.
GrecoFPolliA. Emotional text mining: customer profiling in brand management. Int J Inf Manag2020; 51: 101934.
30.
SatuMSKhanMIMahmudM, et al. TClustVID: a novel machine learning classification model to investigate topics and sentiment in COVID-19 tweets. Knowl-Based Syst2021; 226: 107126.
31.
ZhongJLiuWWangS, et al. Review of methods and applications of text sentiment analysis. Data Anal Knowl Discov2021; 5(6): 1–13.
32.
ZhuXSongJZhangX, et al. Review of text emotion analysis based on topic mining technology. Inf Stud Theory Appl2019; 42(11): 156–163.
33.
LiQWeiCDangJ, et al. Tracking and analyzing public emotion evolutions during COVID-19: a case study from the event-driven perspective on microblogs. Int J Environ Res Public Health2020; 17(18): 6888.
34.
AfrozNBoralMSharmaV, et al. Sentiment analysis of COVID-19 nationwide lockdown effect in India. In: 2021 international conference on artificial intelligence and smart systems (ICAIS), Coimbatore, India, 25–27 March 2021, pp. 561–567. New York: IEEE.
35.
MishraNKSinghPK. Feature construction and smote-based imbalance handling for multi-label learning. Inf Sci2021; 563: 342–357.
36.
WangSWangZJiQ. Multiple emotional tagging of multimedia data by exploiting dependencies among emotions. Multimed Tools Appl2015; 74: 1863–1883.
37.
PereiraRBPlastinoAZadroznyB, et al. Correlation analysis of performance measures for multi-label classification. Inf Process Manag2018; 54(3): 359–369.
SunKWLeeCH. Addressing class-imbalance in multi-label learning via two-stage multi-label hypernetwork. Neurocomputing2017; 266: 375–389.
40.
MoyanoJMGibajaELCiosKJ, et al. Review of ensembles of multi-label classifiers: models, experimental study and prospects. Inf Fusion2018; 44: 33–45.
41.
WinataGIKhodraML. Handling imbalanced dataset in multi-label text categorization using bagging and adaptive boosting. arXiv e-prints arXiv1810, 2018.
42.
WuGTianYLiuD. Cost-sensitive multi-label learning with positive and negative label pairwise correlations. Neural Netw2018; 108: 411–423.
43.
CharteFRiveraAJdel JesusMJ, et al. Addressing imbalance in multilabel classification: measures and random resampling algorithms. Neurocomputing2015; 163: 3–16.
44.
HasaninTKhoshgoftaarT. The effects of random undersampling with simulated class imbalance for big data. In: 2018 IEEE international conference on information reuse and integration (IRI), Salt Lake City, UT, 6–9 July 2018, pp. 70–79. New York: IEEE.
45.
FernándezAGarciaSHerreraF, et al. SMOTE for learning from imbalanced data: progress and challenges, marking the 15-year anniversary. J Artif Intel Res2018; 61: 863–905.
46.
CharteFRiveraAJdel JesusMJ, et al. MLeNN: a first approach to heuristic multilabel undersampling. In: Intelligent data engineering and automated learning–IDEAL 2014: 15th international conference, Salamanca, Spain, 10–12 September 2014, proceedings 15, pp. 1–9. Cham: Springer International Publishing.
47.
PereiraRMCostaYMSillaCNJr.MLTL: a multi-label approach for the Tomek Link undersampling algorithm. Neurocomputing2020; 383: 95–105.
48.
AlmeidaAMCerriRParaisoEC, et al. Applying multi-label techniques in emotion identification of short texts. Neurocomputing2018; 320: 35–46.
49.
JabreelMMorenoA. A deep learning-based approach for multi-label emotion classification in tweets. Appl Sci2019; 9(6): 1123.
50.
ChermanEAMonardMCMetzJ. Multi-label problem transformation methods: a case study. CLEI Electron J2011; 14(1): 1–10.
51.
PushpaMKarpagavalliS. Multi-label classification: problem transformation methods in Tamil phoneme classification. Procedia Comput Sci2017; 115: 572–579.
52.
GasseMAussemAElghazelH. A hybrid algorithm for Bayesian network structure learning with application to multi-label learning. Expert Syst Appl2014; 41(15): 6755–6772.
53.
QianYLiuWHuangJ. A self-attentive convolutional neural networks for emotion classification on user-generated contents. IEEE Access2019; 8: 154198–154208.
54.
QuanCRenF. Automatic annotation of word emotion in sentences based on Ren-CECps. In: Proceedings of the seventh international conference on language resources and evaluation (LREC’10), Valletta, May 2010.
55.
XuLLinHPanY, et al. Constructing the affective lexicon ontology. J China Soc Sci Tech Inf2008; 27(2): 180–185.
PlutchikR. The nature of emotions: human emotions have deep evolutionary roots, a fact that may explain their complexity and provide tools for clinical practice. Am Sci2001; 89(4): 344–350.
58.
SarojaSHaseenaSMadavanR. Dissolved gas analysis of transformer: an approach based on ML and MCDM. IEEE Trans Dielectr Electr Insul2023; 30(5): 2429–2438.
59.
CostantiniGParada-CabaleiroECasaliD, et al. The emotion probe: on the universality of cross-linguistic and cross-gender speech emotion recognition via machine learning. Sensors2022; 22(7): 2461.
60.
PereiraRMSillaCN. Dealing with imbalanceness in hierarchical multi-label datasets using multi-label resampling techniques. In: 2018 IEEE 30th international conference on tools with artificial intelligence (ICTAI), Volos, 5–7 November 2018, pp. 818–824. New York: IEEE.
61.
LinNFuSLinX, et al. Multi-label emotion classification based on adversarial multi-task learning. Inf Process Manag2022; 59(6): 103097.
62.
MikolovTChenKCorradoG, et al. Efficient estimation of word representations in vector space. arXiv preprint arXiv1301.3781, 2013.
63.
LeQMikolovT. Distributed representations of sentences and documents. In: International conference on machine learning, Beijing, China, 21–26 June 2014, pp. 1188–1196. New York: PMLR.
64.
CuiYCheWLiuT, et al. Revisiting pre-trained models for Chinese natural language processing. arXiv preprint arXiv 2004.13922, 2020.
65.
CuiYYangZLiuT. PERT: pre-training BERT with permuted language model. arXiv preprint arXiv 2203.06906, 2022.
66.
CuiYCheWWangS, et al. LERT: a linguistically-motivated pre-trained language model. arXiv preprint arXiv 2211.05344, 2022.
67.
DingEChuXLiuZ, et al. A novel hierarchical adaptive feature fusion method for meta-learning. Appl Sci2022; 12(11): 5458.
68.
LiXYRagaRC. BiLSTM model with attention mechanism for sentiment classification on Chinese mixed text comments. IEEE Access2023; 11: 26199–26210.
69.
CaiRQinBChenY, et al. Sentiment analysis about investors and consumers in energy market based on BERT-BiLSTM. IEEE Access2020; 8: 171408–171415.
70.
ZhuangNYanYChenS, et al. Multi-label learning based deep transfer neural network for facial attribute classification. Pattern Recognit2018; 80: 225–240.
71.
SzymanskiPKajdanowiczT. Scikit-multilearn: a scikit-based Python environment for performing multi-label classification. J Mach Learn Res2019; 20(1): 209–230.
72.
YouLLiuYWangS. Crisis emotion classification combining self-attention and semantic dictionary. Inf Stud Theory Appl2022; 45(5): 189–195.
73.
YangCSongXSongW. SentiBERT: pre-training language model combining sentiment information. J Front Comput Sci Technol2020; 14(9): 1563–1570.
74.
ZhangSLiBYinC. Cross-modal sentiment sensing with visual-augmented representation and diverse decision fusion. Sensors2021; 22(1): 74.
75.
GandhiAAdhvaryuKPoriaS, et al. Multimodal sentiment analysis: a systematic review of history, datasets, multimodal fusion methods, applications, challenges and future directions. Inf Fusion2022; 91: 424–444.
76.
DaiXYZhangLDaiWH, et al. Research on emotional mapping of social networks. Manag Rev2016; 28(8): 79–86.
77.
LiGWangXYeGH. A review of emergency early-warning researches in China. Inf Stud Theory Appl2017; 40(7): 138–144.
