Detecting Information-Hiding Lead Generation Behaviors in Social Q&A Platforms: A Multi-Modal Deep Learning Approach
DOI:
https://doi.org/10.67148/fahss-2026-262Keywords:
Social Q&A platforms; information hiding detection; user behavior modeling; multi-modal learning; anomaly detection; lead generation; platform governance; user experience optimization; deep learning; content moderationAbstract
The proliferation of consultation-oriented social platforms has created unprecedented opportunities for information sharing, yet has also given rise to strategic information-hiding behaviors that undermine user experience and platform integrity. This paper addresses a critical but understudied problem: the automatic detection of "lead generation" posts that deliberately withhold essential information (such as pricing, resources, or contact details) to force users into private communication channels. We propose a novel multi-modal deep learning framework, termed IH-LGDetector, that integrates textual semantic analysis, visual content understanding, and user behavioral sequence modeling to identify such manipulative content. Our approach introduces an Information Deficiency Quantification Index (IDQI) that mathematically measures the degree of information incompleteness in posts, enabling fine-grained analysis of the relationship between information hiding and user engagement patterns. We construct a comprehensive dataset comprising 127,843 posts from a major Chinese social Q&A platform, with meticulous annotation of lead generation behaviors. Extensive experiments demonstrate that our proposed framework achieves 94.7% accuracy and 0.967 AUC-ROC, significantly outperforming baseline methods. Furthermore, correlation analysis reveals a strong positive relationship (r=0.78, p<0.001) between IDQI scores and private message trigger rates, validating our theoretical framework. The proposed system offers immediate deployability for platform governance, with inference latency under 15ms per post, representing a substantial contribution to content ecosystem management and user experience optimization.
References
[1] E. Agichtein, C. Castillo, D. Donato, A. Gionis, and G. Mishne, "Finding high-quality content in social media," in Proc. WSDM, 2008, pp. 183–194.
[2] T. Chen, Y. Zhang, and M. Zhang, "Hierarchical neural networks for answer quality prediction in community question answering," in Proc. ACL, 2022, pp. 4521–4532.
[3] Y. Liu, W. Zhang, and X. Li, "Large-scale analysis of content characteristics on Xiaohongshu platform," Information Processing & Management, vol. 59, no. 3, pp. 102–115, 2022.
[4] H. Wang, J. Liu, and S. Chen, "Content value assessment framework for social media platforms: A case study of Xiaohongshu," IEEE Access, vol. 10, pp. 45678–45692, 2022.
[5] J. Zeng, M. Chen, and L. Wang, "Multi-task learning for content quality assessment in social platforms," in Proc. KDD, 2023, pp. 3345–3355.
[6] A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, et al., "Learning transferable visual models from natural language supervision," in Proc. ICML, 2021, pp. 8748–8763.
[7] W. Kim, B. Son, and I. Kim, "ViLT: Vision-and-language transformer without convolution or region supervision," in Proc. ICML, 2021, pp. 5583–5594.
[8] Y. Zhou, L. Zheng, and C. Wang, "Multi-modal fake news detection via cross-modal attention," in Proc. ACL, 2023, pp. 2341–2352.
[9] X. Wang, Y. Chen, and Z. Liu, "Detecting clickbait posts through multi-modal consistency analysis," IEEE Transactions on Multimedia, vol. 25, pp. 1234–1246, 2023.
[10] J. Huang, W. Li, and M. Zhang, "OCR-enhanced content analysis for Chinese social media platforms," in Proc. EMNLP, 2022, pp. 5678–5689.
[11] J. Fridrich, Steganography in Digital Media: Principles, Algorithms, and Applications. Cambridge University Press, 2009.
[12] A. Westfeld and A. Pfitzmann, "Attacks on steganographic systems," in Proc. Information Hiding Workshop, 1999, pp. 61–76.
[13] F. Wu, J. Huang, and X. Chen, "Adversarial content detection in e-commerce reviews," in Proc. WWW, 2022, pp. 2345–2356.
[14] L. Chen, Y. Wang, and H. Zhang, "Strategic behaviors in social media marketing: Detection and analysis," ACM Transactions on Information Systems, vol. 41, no. 2, pp. 1–28, 2023.
[15] A. Mathur, G. Acar, M. J. Friedman, et al., "Dark patterns at scale: Findings from a crawl of 11K shopping websites," Proc. ACM on Human-Computer Interaction, vol. 3, pp. 1–32, 2019.
[16] B. Hidasi, A. Karatzoglou, L. Baltrunas, and D. Tikk, "Session-based recommendations with recurrent neural networks," in Proc. ICLR, 2016.
[17] W. Kang and J. McAuley, "Self-attentive sequential recommendation," in Proc. ICDM, 2018, pp. 197–206.
[18] D. Wang, J. Lin, and P. Li, "Graph neural networks for fake account detection," in Proc. KDD, 2022, pp. 1987–1997.
[19] Y. Liu, X. Zhang, and W. Chen, "Self-supervised learning for anomalous posting behavior detection," in Proc. NAACL, 2023, pp. 1234–1245.
[20] H. Zhang, Y. Li, and J. Wang, "Information completeness and user engagement: A large-scale analysis," Information Sciences, vol. 612, pp. 456–471, 2022.
[21] F. T. Liu, K. M. Ting, and Z. H. Zhou, "Isolation forest," in Proc. ICDM, 2008, pp. 413–422.
[22] B. Schölkopf, R. C. Williamson, A. J. Smola, J. Shawe-Taylor, and J. C. Platt, "Support vector method for novelty detection," in Proc. NIPS, 2000, pp. 582–588.
[23] M. Sakurada and T. Yairi, "Anomaly detection using autoencoders with nonlinear dimensionality reduction," in Proc. MLSDA Workshop, 2014, pp. 4–11.
[24] L. Zong, B. Chen, and Q. Zhang, "Deep autoencoding Gaussian mixture model for unsupervised anomaly detection," in Proc. ICLR, 2018.
[25] P. Malhotra, A. Ramaswamy, V. Sujatha, et al., "LSTM-based encoder-decoder for multi-sensor anomaly detection," in Proc. ICML Anomaly Detection Workshop, 2016.
[26] Z. Liu, Y. Wang, and X. Chen, "Graph convolutional networks for fraudulent user group detection," in Proc. WWW, 2023, pp. 1567–1578.
[27] S. T. Roberts, Behind the Screen: Content Moderation in the Shadows of Social Media. Yale University Press, 2019.
[28] A. Beutel, J. Chen, T. Doshi, et al., "Fairness in recommendation ranking through pairwise comparisons," in Proc. KDD, 2019, pp. 2212–2220.
[29] J. Wei, Y. Liu, and X. He, "Quality-aware recommendation: Combating low-quality content through algorithmic intervention," in Proc. RecSys, 2023, pp. 234–245.
[30] J. Devlin, M. W. Chang, K. Lee, and K. Toutanova, "BERT: Pre-training of deep bidirectional transformers for language understanding," in Proc. NAACL, 2019, pp. 4171–4186.
[31] A. Dosovitskiy, L. Beyer, A. Kolesnikov, et al., "An image is worth 16x16 words: Transformers for image recognition at scale," in Proc. ICLR, 2021.
[32] Y. Liu, M. Ott, N. Goyal, et al., "RoBERTa: A robustly optimized BERT pretraining approach," arXiv preprint arXiv:1907.11692, 2019.
[33] T. Chen and C. Guestrin, "XGBoost: A scalable tree boosting system," in Proc. KDD, 2016, pp. 785–794.
[34] G. Ke, Q. Meng, T. Finley, et al., "LightGBM: A highly efficient gradient boosting decision tree," in Proc. NIPS, 2017, pp. 3146–3154.
[35] Y. Kim, "Convolutional neural networks for sentence classification," in Proc. EMNLP, 2014, pp. 1746–1751.
[36] Z. Yang, D. Yang, C. Dyer, et al., "Hierarchical attention networks for document classification," in Proc. NAACL, 2016, pp. 1480–1489.
[37] S. M. Lundberg and S. I. Lee, "A unified approach to interpreting model predictions," in Proc. NIPS, 2017, pp. 4765–4774.
[38] M. T. Ribeiro, S. Singh, and C. Guestrin, "'Why should I trust you?': Explaining the predictions of any classifier," in Proc. KDD, 2016, pp. 1135–1144.
[39] T. N. Kipf and M. Welling, "Semi-supervised classification with graph convolutional networks," in Proc. ICLR, 2017.
[40] P. Veličković, G. Cucurull, A. Casanova, et al., "Graph attention networks," in Proc. ICLR, 2018.
[41] A. Vaswani, N. Shazeer, N. Parmar, et al., "Attention is all you need," in Proc. NIPS, 2017, pp. 5998–6008.
[42] I. Loshchilov and F. Hutter, "Decoupled weight decay regularization," in Proc. ICLR, 2019.
[43] H. Touvron, M. Cord, M. Douze, et al., "Training data-efficient image transformers & distillation through attention," in Proc. ICML, 2021, pp. 10347–10357.
[44] J. Li, R. Socher, and S. C. Hoi, "DivideMix: Learning with noisy labels as semi-supervised learning," in Proc. ICLR, 2020.
[45] G. Hinton, O. Vinyals, and J. Dean, "Distilling the knowledge in a neural network," in Proc. NIPS Deep Learning Workshop, 2015.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
For all articles published in Axon Academic Publishing Institute journals, copyright is retained by the authors. Articles are licensed under an open-access Creative Commons CC BY 4.0 license, meaning that anyone may download, read, reuse, and quote the paper for free, provided that the original published version is cited. This ensures that the work receives maximum use and exposure, while giving proper credit to the authors.
In exceptional circumstances, articles may be licensed differently. If you have specific conditions (such as those related to funding) that prevent using this license, please notify the journal’s editorial office during submission. Exceptions will be considered at the publisher's discretion.
