CFTest: Web Based Business Intelligence Application That Measures Crowdfunding Success
No Thumbnail Available
Date
2022
Journal Title
Journal ISSN
Volume Title
Abstract
Crowdfunding (CF), which is implemented and offered to users all over the world as digital entrepreneurship, is a new generation funding type that removes geographical barriers. However, when all CF platforms are considered, the successful project rate on the platforms has entered a decreasing trend as the projects do not comply with a certain standard. To solve this problem, a web-based business intelligence application has been developed to provide decision support to users about their projects by collecting all data from crowdfunding platforms in Turkey using data scraping techniques. The application offered to users under the name of CFTest, uses machine learning methods to deploy projects in the CF ecosystem to a certain standard, predict success, modeling alternative scenarios and visualize summaries, and make a system recommendation to users. Among the models established in this context, the highest performance was achieved with the random forest algorithm, with an accuracy score of 87.52% and an F1 score of 92.16%. Models and plugins transferred to the web environment with the Flask framework are designed in such a way that the user can query. This study presents a model proposal with the application developed by the solution of the problem by considering the decreasing trend in the success rate of crowdfunding platforms. © 2022 IEEE.
Description
Keywords
Crowdsourcing , Decision support systems , Machine learning , User profile , Websites , Business intelligence applications , Business-intelligence , Crowdfunding , Decision supports , Fundings , Machine learning methods , Machine-learning , Performance , Success models , Web-based business , Decision trees