Optimizing B2B customer relationship management and sales forecasting with spectral graph convolutional networks: A quantitative approach

Shagufta Henna; Shyam Krishnan Kalliadan; Mohamed Amjath; Shagufta Henna; Shyam Krishnan Kalliadan; Mohamed Amjath

doi:10.3934/QFE.2025015

Quantitative Finance and Economics

2025, Volume 9, Issue 2: 449-478. doi: 10.3934/QFE.2025015

Previous Article Next Article

Research article

Optimizing B2B customer relationship management and sales forecasting with spectral graph convolutional networks: A quantitative approach

Department of Computing, Atlantic Technological University, Donegal, Ireland

Received: 05 January 2025 Revised: 16 April 2025 Accepted: 16 May 2025 Published: 04 June 2025
JEL Codes: C45, M15

Customer relationship management (CRM) in business-to-business (B2B) environments requires robust strategies and informed decision-making to cultivate strong inter-business relationships, which are pivotal for achieving competitive advantage and maximizing profitability. Traditional CRM analytics, which leverages conventional data mining, machine learning, and deep learning techniques, often fails to address the intricate and interdependent nature of B2B systems. To overcome this limitation, we proposed a spectral graph convolutional neural network (GCN) approach that utilized graph-based modeling to capture the structural complexity of B2B CRM. Companies were represented as nodes, and their interactions as edges within a graph, enriched with Eigenvector centrality and shortest-path graph features, which were particularly suited for spectral GCN operations. Using graph Laplacian-based convolutions, the spectral GCN effectively aggregated global and local relational information, enabling accurate and scalable B2B sales predictions. Experimental evaluations demonstrated that GCN models with spectral attributes significantly outperformed state-of-the-art machine learning and deep learning models, including random forests, convolutional neural networks, feed-forward neural networks, Extreme Gradient Boosting (XGBoost), and Categorical Boosting (CATboost), in terms of accuracy, F1 Score, precision, and specificity. Among the models, the GCN with Eigenvector features achieves the best classification performance, with a high Receiver Operating Characteristic (ROC) value of 0.924, further demonstrating its robustness against variations in feature correlations.
- graph convolutional neural networks,
- business-to-business CRM,
- spectral graph modeling,
- eigenvector centrality,
- sales prediction,
- machine learning analytics
Citation: Shagufta Henna, Shyam Krishnan Kalliadan, Mohamed Amjath. Optimizing B2B customer relationship management and sales forecasting with spectral graph convolutional networks: A quantitative approach[J]. Quantitative Finance and Economics, 2025, 9(2): 449-478. doi: 10.3934/QFE.2025015

Related Papers:

Abstract

Customer relationship management (CRM) in business-to-business (B2B) environments requires robust strategies and informed decision-making to cultivate strong inter-business relationships, which are pivotal for achieving competitive advantage and maximizing profitability. Traditional CRM analytics, which leverages conventional data mining, machine learning, and deep learning techniques, often fails to address the intricate and interdependent nature of B2B systems. To overcome this limitation, we proposed a spectral graph convolutional neural network (GCN) approach that utilized graph-based modeling to capture the structural complexity of B2B CRM. Companies were represented as nodes, and their interactions as edges within a graph, enriched with Eigenvector centrality and shortest-path graph features, which were particularly suited for spectral GCN operations. Using graph Laplacian-based convolutions, the spectral GCN effectively aggregated global and local relational information, enabling accurate and scalable B2B sales predictions. Experimental evaluations demonstrated that GCN models with spectral attributes significantly outperformed state-of-the-art machine learning and deep learning models, including random forests, convolutional neural networks, feed-forward neural networks, Extreme Gradient Boosting (XGBoost), and Categorical Boosting (CATboost), in terms of accuracy, F1 Score, precision, and specificity. Among the models, the GCN with Eigenvector features achieves the best classification performance, with a high Receiver Operating Characteristic (ROC) value of 0.924, further demonstrating its robustness against variations in feature correlations.

References

[1]	Aasman J (2017) Transmuting information to knowledge with an enterprise knowledge graph. IT Professional 19: 44–51. https://doi.org/10.1109/MITP.2017.4241469 doi: 10.1109/MITP.2017.4241469
[2]	Alomrani M, Biparva M, Zhang Y, et al. (2024) DyG2Vec: Efficient Representation Learning for Dynamic Graphs. Trans Mach Learn Res. https://doi.org/10.48550/arXiv.2210.16906
[3]	Asniar and Surendro K (2019) Predictive analytics for predicting customer behavior. 2019 International Conference of Artificial Intelligence and Information Technology (ICAIIT), Yogyakarta, Indonesia, 230–233. https://doi.org/10.1109/ICAIIT.2019.8834571
[4]	Brahim AB, Raboudi W (2020) Improving customer relationship management using machine learning techniques: A Tunisian case study. 2020 International Multi-Conference on: "Organization of Knowledge and Advanced Technologies" (OCTA), Tunis, Tunisia, 1–16. https://doi.org/10.1109/OCTA49274.2020.9151465
[5]	Bruna J, Zaremba W, Szlam A, et al. (2014) Spectral networks and locally connected networks on graphs. Int Conf Learn Representations. https://doi.org/10.48550/arXiv.1312.6203
[6]	B2B Dataset (2020) A real-world dataset for B2B sales analysis. Salvirt Research. Available from: http://www.salvirt.com/research/b2bdataset.
[7]	Cai H, Zheng VW, Chang KC (2018) A comprehensive survey of graph embedding: Problems, techniques, and applications. IEEE Trans Knowl Data Eng 30: 1616–1637. https://doi.org/10.1109/TKDE.2018.2807452 doi: 10.1109/TKDE.2018.2807452
[8]	Chaudhari S, Mithal V, Polatkan G, et al. (2021) An attentive survey of attention models. ACM Trans Intell Syst Technol 12: 1–32. https://doi.org/10.1145/3465055 doi: 10.1145/3465055
[9]	Chung F, Lu L, Vu V (2003) Spectra of random graphs with given expected degrees. Proceedings of the National Acad of Scien 100: 6313–6318. https://doi.org/10.1073/pnas.0937490100 doi: 10.1073/pnas.0937490100
[10]	Das A, Mitra A, Bhagat SN, et al. (2020) Issues and concepts of graph database and a comparative analysis on list of graph database tools. 2020 International Conference on Computer Communication and Informatics (ICCCI), Coimbatore, India, 1–6. https://doi.org/10.1109/ICCCI48352.2020.9104202
[11]	Defferrard M, Bresson X, Vandergheynst P (2016) Convolutional neural networks on graphs with fast localized spectral filtering. Adv Neural Inf Process Sys. https://arXiv.org/abs/1606.09375
[12]	Duvenaud DK, Maclaurin D, Iparraguirre J, et al. (2015) Convolutional networks on graphs for learning molecular fingerprints. Adv Neural Inf Process Syst 28. https://doi.org/10.48550/arXiv.1509.09292 doi: 10.48550/arXiv.1509.09292
[13]	Francis N, Green A, Guagliardo P, et al. (2018) Cypher: An evolving query language for property graphs. Proceedings of the 2018 Int Conf on Manage of Data, 1433–1445. https://doi.org/10.1145/3183713.3190657 doi: 10.1145/3183713.3190657
[14]	Gheisari M, Wang G, Bhuiyan MZA (2017) A survey on deep learning in big data. IEEE Int Conf Comput Sci Eng and IEEE Int Conf Embedded Ubiquitous Comput: 173–180. https://doi.org/10.1109/CSE-EUC.2017.215 doi: 10.1109/CSE-EUC.2017.215
[15]	Global Customer Relationship Management (2020) Growth, Trends and Forecast to 2025. BusinessWire. Available from: https://www.businesswire.com/news/home/20200416005517/en/Global-Customer-Relationship-Management-CRM-Market-Growth.
[16]	Hamilton WL, Ying Z, Leskovec J (2017) Inductive representation learning on large graphs. Adv Neural Inf Process Syst 30. https://doi.org/10.48550/arXiv.1706.02216 doi: 10.48550/arXiv.1706.02216
[17]	Hoksza D, Jelínek J (2015) Using Neo4j for mining protein graphs: A case study. 26th Int Workshop Database Expert Syst Appl, 230–234. https://doi.org/10.1109/DEXA.2015.59 doi: 10.1109/DEXA.2015.59
[18]	Huang H, Dong Z (2013) Research on architecture and query performance based on distributed graph database Neo4j. 3rd Int Conf Consum Electron Commun Netw, 533–536. https://doi.org/10.1109/CECNet.2013.6703387 doi: 10.1109/CECNet.2013.6703387
[19]	Ibrahim A, Ermatita, Saparudin, et al. (2017) Analysis of weakness of data validation from social CRM. Int Conf Data Softw Eng (ICoDSE), 1–5. https://doi.org/10.1109/ICODSE.2017.8285849 doi: 10.1109/ICODSE.2017.8285849
[20]	Kim TS, Lee WK, Sohn SY (2019) Graph convolutional network approach applied to predict hourly bike-sharing demands considering spatial, temporal, and global effects. PLOS ONE 14: e0220782. https://doi.org/10.1371/journal.pone.0220782 doi: 10.1371/journal.pone.0220782
[21]	Kimura D, Gotoh T, Ikeda K (2011) Eliciting considerable requirements with word and customer graphs. IEEE 35th Annu Comput Softw Appl Conf, 476–485. https://doi.org/10.1109/COMPSAC.2011.68 doi: 10.1109/COMPSAC.2011.68
[22]	Kipf TN, Welling M (2017) Semi-supervised classification with graph convolutional networks. Int Conf Learn Representations (ICLR). https://doi.org/10.48550/arXiv.1609.02907 doi: 10.48550/arXiv.1609.02907
[23]	Kitchin R (2014) The data revolution: Big data, open data, data infrastructures & their consequences. SAGE Publications Ltd. https://doi.org/10.4135/9781473909472 doi: 10.4135/9781473909472
[24]	Konno T, Huang R, Ban T, et al. (2017) Goods recommendation based on retail knowledge in a Neo4j graph database combined with an inference mechanism implemented in Jess. IEEE SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI, 1–8. https://doi.org/10.1109/UIC-ATC.2017.8397433 doi: 10.1109/UIC-ATC.2017.8397433
[25]	LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521: 436–444. https://doi.org/10.1038/nature14539 doi: 10.1038/nature14539
[26]	Martínez A, Nin J, Tomás E, et al. (2019) Graph convolutional networks on customer/supplier graph data to improve default prediction. Complex Netw and Their Appl VII, 135–146. https://doi.org/10.1007/978-3-030-14459-3_11 doi: 10.1007/978-3-030-14459-3_11
[27]	Masci J, Boscaini D, Bronstein MM, et al. (2015) Geodesic convolutional neural networks on Riemannian manifolds. IEEE Int Conf Comput Vis Workshop: 832–840. https://doi.org/10.1109/ICCVW.2015.112 doi: 10.1109/ICCVW.2015.112
[28]	McClanahan B, Gokhale SS (2016) Centrality and cluster analysis of Yelp mutual customer business graph. IEEE 40th Annu Comput Softw Appl Conf, 592–601. https://doi.org/10.1109/COMPSAC.2016.79 doi: 10.1109/COMPSAC.2016.79
[29]	Micheli A (2009) Neural network for graphs: A contextual constructive approach. IEEE Trans Neural Networks 20: 498–511. https://doi.org/10.1109/TNN.2008.2010350 doi: 10.1109/TNN.2008.2010350
[30]	Monti F, Boscaini D, Masci J, et al. (2017) Geometric deep learning on graphs and manifolds using mixture model CNNs. 2017 IEEE Conf Comput Vision Pattern Recognit, 5425–5434. https://doi.org/10.1109/CVPR.2017.576 doi: 10.1109/CVPR.2017.576
[31]	Needham M, Hodler AE (2019) Graph Algorithms: Practical Examples in Apache Spark and Neo4j. O'Reilly Media. ISBN: 9781492047681.
[32]	Newman MEJ (2008) Mathematics of networks. In: Durlauf SN, Blume LE (eds) The New Palgrave Dictionary of Economics, 1–8. https://doi.org/10.1057/978-1-349-95121-5_2565-1
[33]	Nugmanova A, Chernykh I, Bulusheva A, et al. (2019) Unsupervised training of automatic dialogue systems for customer support. Int Conf Qual Manag Transp Inf Secur Inf Technol, 436–438. https://doi.org/10.1109/ITQMIS.2019.8928445 doi: 10.1109/ITQMIS.2019.8928445
[34]	Robinson I, Webber J, Eifrem E(2020) Graph Databases, 2nd Edition. Available from: https://www.oreilly.com/library/view/graph-databases-2nd/9781491930885/.
[35]	Saha L, Sahoo L (2018) Adaptation of spectral clustering in telecommunication industry for customer relationship management. 2nd Int Conf Electron Mater Eng Nano-Technol, 1–6. https://doi.org/10.1109/IEMENTECH.2018.8465187 doi: 10.1109/IEMENTECH.2018.8465187
[36]	Scarselli F, Gori M, Tsoi AC, et al. (2009) The graph neural network model. IEEE Trans Neural Networks 20: 61–80. https://doi.org/10.1109/TNN.2008.2005605 doi: 10.1109/TNN.2008.2005605
[37]	Taleb N, Salahat M, Ali L (2020) Impacts of big-data technologies in enhancing CRM performance. 2020 Int Conf Inf Manag, 257–263. https://doi.org/10.1109/ICIM49319.2020.244708 doi: 10.1109/ICIM49319.2020.244708
[38]	Veličković P, Cucurull G, Casanova A, et al. (2018) Graph attention networks. Int Conf Learn Representations. https://doi.org/10.48550/arXiv.1710.10903
[39]	Wang L, Liu S, Pan L, et al. (2014) Enterprise relationship network: Build foundation for social business. IEEE Int Congr Big Data, 347–354. https://doi.org/10.1109/BigData.Congress.2014.57 doi: 10.1109/BigData.Congress.2014.57
[40]	Yang J, Liu C, Teng M, et al. (2018) A unified view of social and temporal modeling for B2B marketing campaign recommendation. IEEE Trans Knowl Data Eng 30: 810–823. https://doi.org/10.1109/TKDE.2017.2783926 doi: 10.1109/TKDE.2017.2783926
[41]	Ye Q, Wang C, Wu B, et al. (2008) VisCRM: A social network visual analytic tool to enhance customer relationship management. IEEE Int Conf Serv Oper Logistics Informatics: 825–830. https://doi.org/10.1109/SOLI.2008.4686513 doi: 10.1109/SOLI.2008.4686513
[42]	Yu A, Dohan D, Le Q, et al. (2018) [Qanet: Combining local convolution with global self-attention for reading comprehension]. Int Conf Learn Representations. https://doi.org/10.48550/arXiv.1804.09541
[43]	Zhang B, Xu D, Zhang H, et al. (2019) STCS lexicon: spectral-clustering-based topic-specific Chinese sentiment lexicon construction for social networks. IEEE Trans Comput Soc Syst 6: 1180–1189. https://doi.org/10.1109/TCSS.2019.2941344 doi: 10.1109/TCSS.2019.2941344
[44]	Zhang Z, Cui P, Zhu W (2022) Deep learning on graphs: A survey. IEEE Trans Knowl Data Eng, 249–270. https://doi.org/10.1109/TKDE.2020.2981333 doi: 10.1109/TKDE.2020.2981333
[45]	Zhao J, Hong Z, Shi M (2019) Analysis of disease data based on Neo4j graph database. IEEE/ACIS 18th Int Conf Comput Inf Sci, 381–384. https://doi.org/10.1109/ICIS46139.2019.8940247 doi: 10.1109/ICIS46139.2019.8940247
[46]	Zhu J, San-Segundo R, Pardo JM (2017) Feature extraction for robust physical activity recognition. Human-Centric Compt and Inf Scien 7: 16. https://doi.org/10.1186/s13673-017-0097-2 doi: 10.1186/s13673-017-0097-2
[47]	Zhuang C, Ma Q (2018) Dual graph convolutional networks for graph-based semi-supervised classification. Proc 2018 World Wide Web Conf, 499–508. https://doi.org/10.1145/3178876.3186116 doi: 10.1145/3178876.3186116
[48]	Zou Y, Liu Y (2020) The implementation knowledge graph of air crash data based on Neo4j. 2020 IEEE 4th Information Technology, Networking, Electronic and Automation Control Conference (ITNEC), Chongqing, China, 1699–1702. https://doi.org/10.1109/ITNEC48623.2020.9085182 doi: 10.1109/ITNEC48623.2020.9085182

Reader Comments

Your name:*

Email:*
© 2025 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)