ANALYSIS OF PROJECT TEXT CONTENTS WITH LSTM IN CROWDFUNDING

Abstract

Crowdfunding (CF) are web platforms that enable projects to be funded and implemented with funding from communities. Thousands of business ideas are successfully implemented with various attributes by using these platforms worldwide every year. One of the attributes that most affect the success of CF is the text content in the projects. In the research conducted in this direction, the summary project texts in the CF platforms operating in Turkey were collected by data scraping techniques and made ready for analysis. Afterwards, the text contents of the CF projects were classified with success tags using an RNN model, LSTM, and analyzed with evaluation metrics. The accuracy rate of the model established with the parameter selections is 96.18%. The results of the study show that the texts prepared for CF projects can be tested in decision support systems.

Keywords

• Crowdfunding
• text mining
• neural networks
• deep learning
• long-short term memory

KİTLE FONLAMASINDAKİ PROJE METİN İÇERİKLERİNİN LSTM İLE ANALİZİ

Abstract

Kitle fonlaması (KF), topluluklardan gelen fonlamalarla projelerin finanse edilerek hayata geçmesini sağlayan web platformlarıdır. Dünya çapında her yıl bu platformlar kullanılarak binlerce iş fikri çeşitli öznitelikler ile başarılı bir şekilde gerçekleştirilmektedir. KF başarısına en çok etki eden özniteliklerden birisi de projelerdeki metin içerikleridir. Bu doğrultuda yapılan araştırmada, Türkiye’de faaliyet gösteren KF platformlarındaki özetleyici proje metinleri veri kazıma teknikleriyle toplanmış ve analize hazır hale getirilmiştir. Sonrasında ise KF projelerinin metin içerikleri bir RNN modeli olan LSTM kullanılarak başarı etiketleriyle sınıflandırılmış ve değerlendirme metrikleriyle analiz edilmiştir. Parametre seçimleriyle birlikte kurulan modelin doğruluk oranı %96.18’dir. Çalışmanın sonuçları, KF projeleri için hazırlanan metinlerin karar destek sistemlerinde test edilebileceğini göstermektedir.

Keywords

• kitle Fonlaması
• metin madenciliği
• sinir ağları
• derin öğrenme
• uzun-kısa süreli bellek

References

1. Akça, M. F. (2021). LSTM Nedir? Nasıl Çalışır? Erişim Tarihi: 12.07.2021, Erişim Linki: https://mfakca.medium.com/lstm-nedir-nasıl-çalışır-326866fd8869
2. Akdoğan, A. (2020). Uzun Kısa Vadeli Hafıza Ağları. Erişim Tarihi: 18.07.2021, Erişim Linki: https://medium.com/bilişim-hareketi/uzun-kısa-vadeli-hafıza-ağları-lstm-95cbe7d51b44
3. Akköse, O. (2020). Uzun-Kısa Vadeli Bellek (LSTM). Erişim Tarihi: 12.07.2021, Erişim Linki: https://medium.com/deep-learning-turkiye/uzun-kısa-vadeli-bellek-lstm-b018c07174a3
4. Basiri, M. E., Nemati, S., Abdar, M., Cambria, E., & Acharya, U. R. (2021). ABCDM: An Attention-based Bidirectional CNN-RNN Deep Model for sentiment analysis. Future Generation Computer Systems, 115, 279–294. https://doi.org/10.1016/j.future.2020.08.005
5. Bilgin, M., & Şentürk, İ. F. (2017). Sentiment analysis on Twitter data with semi-supervised Doc2Vec. 2nd International Conference on Computer Science and Engineering, UBMK 2017, 661–666. https://doi.org/10.1109/UBMK.2017.8093492
6. Borrero-Domínguez, C., Cordón-Lagares, E., & Hernández-Garrido, R. (2020). Analysis of success factors in crowdfunding projects based on rewards: A way to obtain financing for socially committed projects. Heliyon, 6(4). https://doi.org/10.1016/j.heliyon.2020.e03744
7. Chakraborty, S., & Swinney, R. (2020). Signaling to the Crowd : Private Quality Information and Rewards-Based Crowfunding. Manufacturing & Service Operations Management, April, 0–15.
8. Elnagar, A., Al-Debsi, R., & Einea, O. (2020). Arabic text classification using deep learning models. Information Processing and Management, 57(1), 102121. https://doi.org/10.1016/j.ipm.2019.102121

9. Farhoud, M., Shah, S., Stenholm, P., Kibler, E., Renko, M., & Terjesen, S. (2021). Social enterprise crowdfunding in an acute crisis. Journal of Business Venturing Insights, 15(November 2020), e00211. https://doi.org/10.1016/j.jbvi.2020.e00211
10. Hu, J., Wang, X., Zhang, Y., Zhang, D., Zhang, M., & Xue, J. (2020). Time Series Prediction Method Based on Variant LSTM Recurrent Neural Network. Neural Processing Letters, 52(2), 1485–1500. https://doi.org/10.1007/s11063-020-10319-3
11. Jang, B., Kim, M., Harerimana, G., Kang, S. U., & Kim, J. W. (2020). Bi-LSTM model to increase accuracy in text classification: Combining word2vec CNN and attention mechanism. Applied Sciences (Switzerland), 10(17). https://doi.org/10.3390/app10175841
12. Kızrak, M. A., & Bolat, B. (2019). Uçak Motoru Sağlığı için Uzun-Kısa Süreli Bellek Yöntemi ile Öngörücü Bakım. Bilişim Teknolojileri Dergisi, 103–109. https://doi.org/10.17671/gazibtd.495730
13. Li, Yan, Rakesh, V., & Reddy, C. K. (2016). Project success prediction in crowdfunding environments. WSDM 2016 - Proceedings of the 9th ACM International Conference on Web Search and Data Mining, 247–256. https://doi.org/10.1145/2835776.2835791
14. Li, Yue, Wang, X., & Xu, P. (2018). Chinese text classification model based on deep learning. Future Internet, 10(11). https://doi.org/10.3390/fi10110113
15. Lukkarinen, A., Teich, J. E., Wallenius, H., & Wallenius, J. (2016). Success drivers of online equity crowdfunding campaigns. Decision Support Systems, 87, 26–38. https://doi.org/10.1016/j.dss.2016.04.006
16. Matsubara, N., Teramoto, A., Saito, K., & Fujita, H. (2019). Generation of Pseudo Chest X-ray Images from Computed Tomographic Images by Nonlinear Transformation and Bone Enhancement. Medical Imaging and Information Sciences, 36(3), 141–146. https://doi.org/10.11318/mii.36.141
17. Moradi, M., & Badrinarayanan, V. (2021). The effects of brand prominence and narrative features on crowdfunding success for entrepreneurial aftermarket enterprises. Journal of Business Research, 124(November 2020), 286–298. https://doi.org/10.1016/j.jbusres.2020.12.002
18. Nergiz, G., Safali, Y., Avaroglu, E., & Erdogan, S. (2019). Classification of Turkish News Content by Deep Learning Based LSTM Using Fasttext Model. 2019 International Conference on Artificial Intelligence and Data Processing Symposium, IDAP 2019, 1–6. https://doi.org/10.1109/IDAP.2019.8875949
19. Ryoba, M. J., Qu, S., & Zhou, Y. (2020). Feature subset selection for predicting the success of crowdfunding project campaigns. Electronic Markets, 1–14. https://doi.org/10.1007/s12525-020-00398-4
20. Seyyarer, E., Ayata, F., Uçkan, T., & Karcı, A. (2020). Derin Öğrenmede Kullanılan Optimizasyon Algoritmalarının Uygulanması Ve Kıyaslanması. Anatolian Journal of Computer Sciences, 2, 90–98.
21. Shneor, R., & Vik, A. A. (2020). Crowdfunding success: a systematic literature review 2010–2017. In Baltic Journal of Management (Vol. 15, Issue 2, pp. 149–182). Emerald Group Publishing Ltd. https://doi.org/10.1108/BJM-04-2019-0148
22. Zhou, C., Sun, C., Liu, Z., & Lau, F. C. M. (2015). A C-LSTM Neural Network for Text Classification. http://arxiv.org/abs/1511.08630