Kitle Fonlaması Projelerinin Karar Ağacı ve Rastgele Orman Algoritmalarıyla Sınıflandırılması

Öz

Kitle fonlaması platformları, internet üzerinden iş fikirlerini hayata geçirme ya da destek alabilme noktasında büyük olanaklar sağlayabilmektedir. Bu platformlarda destek beklenen projelerin başarısı, alınan finansal destek ile doğru orantılı bir şekilde artmaktadır. Fakat finansal destek alabilmek için projenin destekçilere iyi bir şekilde sunulması gerekir. Günümüzde bu platformlar iyi tasarlanmamış projelerle dolu olduğu için başarı oranı oldukça düşüş göstermiştir. Bu sebeple, finansal destek alınabilmesi için projelerin başarı anlamında test edilmesi ve başarısız olarak sınıflandırılan projelerin eksiklerini gidererek destekçilere yeniden sunulması gerekmektedir. Bu kapsamda, ortaya koyduğumuz çalışmada birçok kategorideki Kickstarter projesi makine öğrenmesi yöntemleriyle sınıflandırılarak web arayüzünde son kullanıcıya sunulmuştur. Projelerin sınıflandırılması için, dağınık veri setlerinde iyi sınıflandırma yapabilen Decision Tree ve Random Forest algoritmaları kullanılmıştır. Algoritmalar, sırasıyla %73 ve %81 oranında sınıflandırma yapabilmektedir. Ayrıca, yapılan sınıflandırmalar değerlendirme metrikleriyle de test edilerek ne kadar doğru sınıflandırma yapılabildiği ölçülmüştür. Bu sayede, kitle fonlaması platformlarına projelerini ekleyen veya ekleyecek olan girişimciler, finansal bir destek aramadan önce projelerini başarı anlamında test ederek eksiklerini görebileceklerdir.

Anahtar Kelimeler

• Makine Öğrenmesi
• Web Uygulamaları
• Yönetim Bilişim Sistemleri

Classification of Crowdfunding Projects by Decision Tree and Random Forest Algorithms

Öz

Crowdfunding platforms can provide great opportunities to implement business ideas or get support over the internet. The success of projects that are expected to support these platforms increases in direct proportion to the financial support received. But in order to receive financial support, the project must be well presented to backers. Today, the success rate has declined considerably because these platforms are full of poorly designed projects. For this reason, in order to receive financial support, projects must be tested in terms of success and re-presented to supporters by eliminating the deficiencies of projects classified as unsuccessful. In this context, in our study, many categories of Kickstarter projects are classified by machine learning methods and presented to the end user in the web interface. For the classification of projects, decision Tree and Random Forest algorithms that can classify well in scattered data sets were used. Algorithms can classify by 73% and 81%, respectively. In addition, the classifications made were also tested
with evaluation metrics and measured how accurate the classification can be made. In this way,
entrepreneurs who add or will add their projects to crowdfunding platforms will be able to see
their shortcomings by testing their projects for success before receiving financial support.

Kaynakça

1. [1] Laudon, K. C. (2007). Management information systems: Managing the digital firm. Pearson Education India.
2. [2] Theis, T. N., & Wong, H. S. P. (2017). The end of moore's law: A new beginning for information technology. Computing in Science & Engineering, 19(2), 41.
3. [3] Mansoori, Y., Karlsson, T., & Lundqvist, M. (2019). The influence of the lean startup methodology on entrepreneur-coach relationships in the context of a startup accelerator. Technovation, 84, 37-47.
4. [4] Zvilichovsky, D., Inbar, Y., & Barzilay, O. (2015). Playing both sides of the market: Success and reciprocity on crowdfunding platforms. Available at SSRN 2304101.
5. [5] Etter, V., Grossglauser, M., & Thiran, P. (2013, October). Launch hard or go home! Predicting the success of Kickstarter campaigns. In Proceedings of the first ACM conference on Online social networks (pp. 177-182).
6. [6] Kuppuswamy, V., & Bayus, B. L. (2018). Crowdfunding creative ideas: The dynamics of project backers. In The economics of crowdfunding (pp. 151-182). Palgrave Macmillan, Cham.
7. [7] Cheng, C., Tan, F., Hou, X., & Wei, Z. (2019, August). Success prediction on crowdfunding with multimodal deep learning. In Proceedings of the 28th International Joint Conference on Artificial Intelligence, Macao, China (pp. 10-16).
8. [8] Chen, K., Jones, B., Kim, I., & Schlamp, B. (2013). Kickpredict: Predicting Kickstarter Success. Technical report, California Institute of Technology.

9. [9] Kindler, A., Golosovsky, M., & Solomon, S. (2019). Early Prediction of the Outcome of Kickstarter Campaigns: Is the Success due to Virality? Palgrave Communications, 5(1), 1-6.
10. [10] Chung, J., & Lee, K. (2015, August). A long-term study of a crowdfunding platform: Predicting project success and fundraising amount. In Proceedings of the 26th ACM Conference on Hypertext & Social Media (pp. 211-220).
11. [11] Rao, H., Xu, A., Yang, X., & Fu, W. T. (2014, April). Emerging dynamics in crowdfunding campaigns. In International Conference on Social Computing, Behavioral-Cultural Modeling, and Prediction (pp. 333-340). Springer, Cham.
12. [12] Jensen, L. S., & Özkil, A. G. (2018). Identifying challenges in crowdfunded product development: a review of Kickstarter projects. Design Science, 4.
13. [13] Du, Q., Fan, W., Qiao, Z., Wang, G., Zhang, X., & Zhou, M. (2015). Money talks: a predictive model on crowdfunding success using project description.
14. [14] Bi, S., Liu, Z., & Usman, K. (2017). The influence of online information on investing decisions of reward-based crowdfunding. Journal of Business Research, 71, 10-18.
15. [15] Mortensen, S., Christison, M., Li, B., Z hu, A., & Venkatesan, R. (2019, April). Predicting and Defining B2B Sales Success with Machine Learning. In 2019 Systems and Information Engineering Design Symposium (SIEDS) (pp. 1-5). IEEE.
16. [16] Mouillé, M. (2018). Kickstarter Projects Dataset, Kaggle. More than 300,000 kickstarter projects (Version 7). Access address: https://www.kaggle.com/kemical/kickstarter-projects
17. [17] Huber, S., Wiemer, H., Schneider, D., & Ihlenfeldt, S. (2019). DMME: Data mining methodology for engineering applications–a holistic extension to the CRISP-DM model. Procedia CIRP, 79, 403-408.
18. [18] Fahmy, A. F., Mohamed, H. K., & Yousef, A. H. (2017). A data mining experimentation framework to improve six sigma projects. In 2017 13th International Computer Engineering Conference (ICENCO) (pp. 243-249). IEEE.
19. [19] Berhane, T. M., Lane, C. R., Wu, Q., Autrey, B. C., Anenkhonov, O. A., Chepinoga, V. V., & Liu, H. (2018). Decision-tree, rule-based, and random forest classification of high-resolution multispectral imagery for wetland mapping and inventory. Remote sensing, 10(4), 580.
20. [20] Leiva, R. G., Anta, A. F., Mancuso, V., & Casari, P. (2019). A Novel Hyperparameter-Free Approach to Decision Tree Construction That Avoids Overfitting by Design. IEEE Access, 7, 99978-99987.
21. [21] Pal, M. (2005). Random forest classifier for remote sensing classification. International Journal of Remote Sensing, 26(1), 217-222.
22. [22] Xu, J., Zhang, Y., & Miao, D. (2020). Three-way confusion matrix for classification: a measure driven view. Information Sciences, 507, 772-794.
23. [23] Yang, T. L., Lin, C. H., Chen, W. L., Lin, H. Y., Su, C. S., & Liang, C. K. (2019). Hash Transformation and Machine Learning-based Decision-Making Classifier Improved the Accuracy Rate of Automated Parkinson’s Disease Screening. IEEE Transactions on Neural Systems and Rehabilitation Engineering.
24. [24] Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and psychological measurement, 20(1), 37-46.
25. [25] Wang, J., Wu, X., & Zhang, C. (2005). Support vector machines based on K-means clustering for real-time business intelligence systems. International Journal of Business Intelligence and Data Mining, 1(1), 54-64.
26. [26] Cook, A., Wu, P., & Mengersen, K. (2015). Machine learning and visual analytics for consulting business decision support. In 2015 Big Data Visual Analytics (BDVA) (pp. 1-2). IEEE.