Abstract
Based on the product review corpus, it is difficult to accurately analyze the emotional polarity of candidate emotion words without considering the domain-oriented emotional polarity weak tags. In order to solve this problem, this paper proposes a method for calculating the weak tags of word sentiment polarity based on product attributes and relevance. Based on the obtained subject and attributes of the goods, the user’s satisfaction with them and the relevance between them and the candidate emotion words are calculated by using enhanced point mutual information. Finally, the value of the emotional polarity weak label of the candidate emotional word is calculated according to the satisfaction degree and the degree of association. Experimental results show that calculating the emotional polarity labels of candidate emotion words can effectively improve the accuracy of emotion classification.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Miquel, P.N., Raul, S.R., Dario, G.G., Jesus, C.S.: Recycling weak labels for multiclass classification. Neurocomputing (2020)
Jinyuan, W.U., Junzhong, J.I., Xuewu, Z., Chensheng, W.U., Fanghua, D.U.: Weight calculation of emotional word based on feature selection technique. J. Beijing Univ. Technol. 42, 142–151 (2016)
Duan, J., Li, X., Mu, D.: Learning multi labels from single label—an extreme weak label learning algorithm. Wuhan University J. Nat. Sci. 24(2), 161–168 (2019)
Zhang, S., He, L., Vucetic, S., Dragut, E.C.: Regular expression guided entity mention mining from noisy web data. In: EMNLP, pp. 1–9 (2018)
Miquel, P.N., Raúl, S.R., Jesús, C.S.: Adapting supervised classification algorithms to arbitrary weak label scenarios, pp. 247–259 (2017)
Adavanne, S., Virtanen, T.: Sound event detection using weakly labeled dataset with stacked convolutional and recurrent neural network, pp. 1–5 (2017)
Olivier, J., Steven, V., Erik, M., Sofie, V.H., Rik, V.W.: Influence of weak labels for emotion recognition of tweets. Mining Intell. Knowl. Explor., 108–118 (2014)
Acknowledge
This Research work was supported in part by 2018 Cultivation Project of Top Talent in Anhui Colleges and Universities (Grant No. gxbjZD15), in part by 2019 Anhui Pro-vincial Natural Science Foundation Project (Grant No. 1908085MF189).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Yu, H., Zhang, S. (2021). Constructing Emotional Weak Labels for Online Shopping Platform Based on Product Attribute and Relevance. In: Abawajy, J., Choo, KK., Xu, Z., Atiquzzaman, M. (eds) 2020 International Conference on Applications and Techniques in Cyber Intelligence. ATCI 2020. Advances in Intelligent Systems and Computing, vol 1244. Springer, Cham. https://doi.org/10.1007/978-3-030-53980-1_19
Download citation
DOI: https://doi.org/10.1007/978-3-030-53980-1_19
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-53979-5
Online ISBN: 978-3-030-53980-1
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)